{ "generatedAt": "2026-06-12T23:06:14.466Z", "branch": "staging", "items": [ { "id": "doc:docs/runbooks/2026-05-22-prod-cleanup.md", "kind": "doc", "title": "2026-05-22 Prod Cleanup Runbook", "summary": "Loose ends from the 2026-05-22 prod fire. These are ops actions a human\nneeds to perform; nothing in this list is automatable from the code path.\nThe URL `postgresql://postgres:@mainline.proxy.rlwy.net:44431/railway`\nleaked in `docs/superpowers/plans/*.md` on staging commits b58fabac and", "path": "docs/runbooks/2026-05-22-prod-cleanup.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/runbooks/2026-05-22-prod-cleanup.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "runbooks" ], "searchText": "2026-05-22 prod cleanup runbook loose ends from the 2026-05-22 prod fire. these are ops actions a human\nneeds to perform; nothing in this list is automatable from the code path.\nthe url `postgresql://postgres:@mainline.proxy.rlwy.net:44431/railway`\nleaked in `docs/superpowers/plans/*.md` on staging commits b58fabac and docs/runbooks/2026-05-22-prod-cleanup.md https://github.com/torus-intelligence/torus/blob/staging/docs/runbooks/2026-05-22-prod-cleanup.md github runbook active docs runbook docs runbooks" }, { "id": "doc:docs/superpowers/audits/2026-05-28-staging-full.md", "kind": "doc", "title": "2026-05-28 Staging Full Audit Sweep", "summary": "**Branch:** `cleanup/staging-audit-sweep` from `origin/staging`\n**Scope:** full repo pass based on the 2026-04-21 mega-audit design and the\n2026-05-20/21 weekly audit threads. Seven pods re-ran across auth/session,\nagent/runtime, process/change-order, graph/RLS/schema, performance/storage,", "path": "docs/superpowers/audits/2026-05-28-staging-full.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-28-staging-full.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits" ], "searchText": "2026-05-28 staging full audit sweep **branch:** `cleanup/staging-audit-sweep` from `origin/staging`\n**scope:** full repo pass based on the 2026-04-21 mega-audit design and the\n2026-05-20/21 weekly audit threads. seven pods re-ran across auth/session,\nagent/runtime, process/change-order, graph/rls/schema, performance/storage, docs/superpowers/audits/2026-05-28-staging-full.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-28-staging-full.md github runbook active docs runbook docs superpowers audits" }, { "id": "skill:account-research", "kind": "autopilot-skill", "title": "Account Research", "summary": "---\nname: account-research\ndescription: |\nUse when researching a target EPC firm, preparing for a call/meeting, building a positioning argument, handling an objection, or cutting a vertical brief.", "path": "ops/autopilot/skills/account-research/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/account-research/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "account-research" ], "searchText": "account research ---\nname: account-research\ndescription: |\nuse when researching a target epc firm, preparing for a call/meeting, building a positioning argument, handling an objection, or cutting a vertical brief. ops/autopilot/skills/account-research/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/account-research/skill.md autopilot active autopilot skill skills account-research" }, { "id": "doc:docs/designs/agent-memory-architecture.md", "kind": "doc", "title": "Agent Memory Architecture", "summary": "---\nstatus: APPROVED\nbase: post-narrative-claims (Phase 1 + Phase 2 of 2026-05-20-narrative-claims-megaplan.md)\n---", "path": "docs/designs/agent-memory-architecture.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/designs/agent-memory-architecture.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "designs" ], "searchText": "agent memory architecture ---\nstatus: approved\nbase: post-narrative-claims (phase 1 + phase 2 of 2026-05-20-narrative-claims-megaplan.md)\n--- docs/designs/agent-memory-architecture.md https://github.com/torus-intelligence/torus/blob/staging/docs/designs/agent-memory-architecture.md github architecture active docs architecture docs designs" }, { "id": "runner:agent-output-analyzer", "kind": "autopilot-runner", "title": "Agent Output Analyzer", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/agent-output-analyzer/run.ts.", "path": "ops/autopilot/runners/agent-output-analyzer/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/agent-output-analyzer/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "agent-output-analyzer" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "agent output analyzer scheduled or operator-run autopilot workflow at ops/autopilot/runners/agent-output-analyzer/run.ts. ops/autopilot/runners/agent-output-analyzer/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/agent-output-analyzer/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner agent-output-analyzer" }, { "id": "skill:agent-output-analyzer", "kind": "autopilot-skill", "title": "Agent Output Analyzer", "summary": "---\nname: agent-output-analyzer\ndescription: |\nWeekly analysis of which agent plans get approved unchanged, edited, or rejected.", "path": "ops/autopilot/skills/agent-output-analyzer/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/agent-output-analyzer/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "agent-output-analyzer" ], "searchText": "agent output analyzer ---\nname: agent-output-analyzer\ndescription: |\nweekly analysis of which agent plans get approved unchanged, edited, or rejected. ops/autopilot/skills/agent-output-analyzer/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/agent-output-analyzer/skill.md autopilot active autopilot skill skills agent-output-analyzer" }, { "id": "module:agent-query", "kind": "module", "title": "Agent Query Boundary", "summary": "Planned Ops API/MCP bridge for Claude, Codex, Granola, CRM, Railway, GitHub docs, and Autopilot state.", "href": "#agent", "source": "OPS", "status": "planned", "tags": [ "mcp", "codex", "claude", "api" ], "nextAction": "Build ops.search, ops.read_doc, ops.search_crm, and append-only CRM write tools.", "searchText": "agent query boundary planned ops api/mcp bridge for claude, codex, granola, crm, railway, github docs, and autopilot state. #agent ops planned build ops.search, ops.read_doc, ops.search_crm, and append-only crm write tools. mcp codex claude api" }, { "id": "doc:docs/superpowers/plans/2026-05-27-approvals-spine.SUPERSEDED.md", "kind": "doc", "title": "Approvals Spine Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Build the document-approval primitive that unifies change notices (with a process-engineering FDCN flavor), client squad checks, and future approval surfaces into a single typed state machine with token-based external recipients.\n**Architecture:** New `backend/plugins/approvals/` plugin. Generic `change_notice` flow kind with discipline-specific flavors (process-engineering FDCN, piping, mechanical, electrical, instrumentation, civil). A second `squad_check` flow kind for client review cycles. Request-level state separate from per-step state. External reviewer...", "path": "docs/superpowers/plans/2026-05-27-approvals-spine.SUPERSEDED.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-27-approvals-spine.SUPERSEDED.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "approvals spine implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** build the document-approval primitive that unifies change notices (with a process-engineering fdcn flavor), client squad checks, and future approval surfaces into a single typed state machine with token-based external recipients.\n**architecture:** new `backend/plugins/approvals/` plugin. generic `change_notice` flow kind with discipline-specific flavors (process-engineering fdcn, piping, mechanical, electrical, instrumentation, civil). a second `squad_check` flow kind for client review cycles. request-level state separate from per-step state. external reviewer... docs/superpowers/plans/2026-05-27-approvals-spine.superseded.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-27-approvals-spine.superseded.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:ops/autopilot/context/org/architecture.md", "kind": "doc", "title": "Architecture Overview", "summary": "| Service | Stack | Where it runs | Database | Notes |\n|---------|-------|---------------|----------|-------|\n| backend | Bun + TypeScript + plugin system | Railway | Postgres 18 + pgvector + pg_trgm | Plugin-based, RLS-scoped queries |\n| sidecar | Python 3.13 + GLiNER2-large + GLiDRE + Docling + bge-small | Railway | -- | ML extraction; vendored models |", "path": "ops/autopilot/context/org/architecture.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/org/architecture.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context", "org" ], "searchText": "architecture overview | service | stack | where it runs | database | notes |\n|---------|-------|---------------|----------|-------|\n| backend | bun + typescript + plugin system | railway | postgres 18 + pgvector + pg_trgm | plugin-based, rls-scoped queries |\n| sidecar | python 3.13 + gliner2-large + glidre + docling + bge-small | railway | -- | ml extraction; vendored models | ops/autopilot/context/org/architecture.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/org/architecture.md autopilot autopilot active docs autopilot context org" }, { "id": "doc:docs/superpowers/audits/INDEX.md", "kind": "doc", "title": "Audits Index", "summary": "One-line summary of every file in `docs/superpowers/audits/`. Per-run subdirectories hold pod reports and per-cluster execution logs.\n- [2026-05-20-weekly.md](./2026-05-20-weekly.md) — v2 deep pass of the weekly mega-audit (306 raw, 246 distinct findings across 7 pods including a two-round adversarial pod).\n- [2026-05-21-weekly.md](./2026-05-21-weekly.md) — Save-point audit re-verifying PR #44 (Wave 0-4 cleanup) against staging tip `8125d2f1`; ~75% of prior findings closed at the cited file:line.\n- [2026-05-28-staging-full.md](./2026-05-28-staging-full.md) — Full staging sweep across all seven pods, cleanup landed in `cleanup/staging-audit-sweep`, and remaining high-risk follow-ups.", "path": "docs/superpowers/audits/INDEX.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/INDEX.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits" ], "searchText": "audits index one-line summary of every file in `docs/superpowers/audits/`. per-run subdirectories hold pod reports and per-cluster execution logs.\n- [2026-05-20-weekly.md](./2026-05-20-weekly.md) — v2 deep pass of the weekly mega-audit (306 raw, 246 distinct findings across 7 pods including a two-round adversarial pod).\n- [2026-05-21-weekly.md](./2026-05-21-weekly.md) — save-point audit re-verifying pr #44 (wave 0-4 cleanup) against staging tip `8125d2f1`; ~75% of prior findings closed at the cited file:line.\n- [2026-05-28-staging-full.md](./2026-05-28-staging-full.md) — full staging sweep across all seven pods, cleanup landed in `cleanup/staging-audit-sweep`, and remaining high-risk follow-ups. docs/superpowers/audits/index.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/index.md github runbook active docs runbook docs superpowers audits" }, { "id": "module:autopilot", "kind": "module", "title": "Autopilot", "summary": "Scheduled runners, skills, inventory, research outputs, sourcing artifacts, and review queues.", "href": "#autopilot", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "agents", "runs" ], "searchText": "autopilot scheduled runners, skills, inventory, research outputs, sourcing artifacts, and review queues. #autopilot autopilot active autopilot agents runs" }, { "id": "doc:ops/autopilot/docs/conventions.md", "kind": "doc", "title": "Autopilot Conventions", "summary": "These are contracts every scheduled skill honors. They are the substrate for quiet-agent posting, the daily digest, and the cross-skill feed graph.\n- `DONE` -- normal success with output\n- `DONE_WITH_CONCERNS` -- success, but flagged items need human review\n- `DONE_EMPTY` -- ran cleanly, nothing to do", "path": "ops/autopilot/docs/conventions.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/conventions.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "autopilot conventions these are contracts every scheduled skill honors. they are the substrate for quiet-agent posting, the daily digest, and the cross-skill feed graph.\n- `done` -- normal success with output\n- `done_with_concerns` -- success, but flagged items need human review\n- `done_empty` -- ran cleanly, nothing to do ops/autopilot/docs/conventions.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/conventions.md autopilot autopilot active docs autopilot docs" }, { "id": "doc:ops/autopilot/docs/setup-guide.md", "kind": "doc", "title": "Autopilot Setup Guide", "summary": "End-to-end guide for provisioning credentials and wiring the first Codex environment. Pair this with `codex-runtime.md` (the runtime model) and the per-agent docs in `ops/autopilot/codex-environments/`.\n---\nSix external systems, in roughly this order:\n1. **Linear** -- API key + `autopilot-reviewed` label UUID + default team ID", "path": "ops/autopilot/docs/setup-guide.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/setup-guide.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "autopilot setup guide end-to-end guide for provisioning credentials and wiring the first codex environment. pair this with `codex-runtime.md` (the runtime model) and the per-agent docs in `ops/autopilot/codex-environments/`.\n---\nsix external systems, in roughly this order:\n1. **linear** -- api key + `autopilot-reviewed` label uuid + default team id ops/autopilot/docs/setup-guide.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/setup-guide.md autopilot autopilot active docs autopilot docs" }, { "id": "skill:blog-autopilot", "kind": "autopilot-skill", "title": "Blog Autopilot", "summary": "---\nname: blog-autopilot\ndescription: |\nScheduled topic discovery + MDX draft + PR open for the Torus blog. Hard dependency on Phase 5 (website blog system).", "path": "ops/autopilot/skills/blog-autopilot/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/blog-autopilot/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "blog-autopilot" ], "searchText": "blog autopilot ---\nname: blog-autopilot\ndescription: |\nscheduled topic discovery + mdx draft + pr open for the torus blog. hard dependency on phase 5 (website blog system). ops/autopilot/skills/blog-autopilot/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/blog-autopilot/skill.md autopilot active autopilot skill skills blog-autopilot" }, { "id": "doc:ops/autopilot/context/brain-first.md", "kind": "doc", "title": "Brain-First Convention (Autopilot)", "summary": "Before hitting external APIs (GitHub, Linear, Discord, Azure, Google),\ncheck the inventory. `inventory/` is autopilot's brain.\n1. **Knowledge**: `inventory/knowledge/` for matching entries\n(patterns, decisions, recurring issues).", "path": "ops/autopilot/context/brain-first.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/brain-first.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "brain-first convention (autopilot) before hitting external apis (github, linear, discord, azure, google),\ncheck the inventory. `inventory/` is autopilot's brain.\n1. **knowledge**: `inventory/knowledge/` for matching entries\n(patterns, decisions, recurring issues). ops/autopilot/context/brain-first.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/brain-first.md autopilot autopilot active docs autopilot context" }, { "id": "doc:ops/autopilot/docs/building-skills.md", "kind": "doc", "title": "Building Autopilot Skills", "summary": "How to create a new agent skill for the Torus autopilot system.\nA skill is a self-contained agent definition that lives in `skills//`. Each skill has a `SKILL.md` file with YAML frontmatter and a markdown body describing the agent's workflow. Skills run as Claude Code scheduled agents in Anthropic's cloud -- they clone this repo on every run and follow the instructions in their SKILL.md.\nEvery skill must have YAML frontmatter with `name` and `description`. The description must start with \"Use when\".\nA typical skill directory:", "path": "ops/autopilot/docs/building-skills.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/building-skills.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "building autopilot skills how to create a new agent skill for the torus autopilot system.\na skill is a self-contained agent definition that lives in `skills//`. each skill has a `skill.md` file with yaml frontmatter and a markdown body describing the agent's workflow. skills run as claude code scheduled agents in anthropic's cloud -- they clone this repo on every run and follow the instructions in their skill.md.\nevery skill must have yaml frontmatter with `name` and `description`. the description must start with \"use when\".\na typical skill directory: ops/autopilot/docs/building-skills.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/building-skills.md autopilot autopilot active docs autopilot docs" }, { "id": "autopilot-output:ops/autopilot/inventory/agent-state/care-friction-logs.json", "kind": "autopilot-output", "title": "Care Friction Logs", "summary": "{\n \"last_run\": \"2026-04-03T09:00:00-06:00\",\n \"last_status\": \"BLOCKED\",\n \"last_blocker\": \"Discord webhook delivery issue (cannot read #dev channel); gws CLI not found (cannot read CARE sheet)\",\n \"last_blocker_posted\": \"2026-04-03\",\n \"last_blocker_first_seen\": \"2026-04-03\"\n}", "path": "ops/autopilot/inventory/agent-state/care-friction-logs.json", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/care-friction-logs.json", "source": "AUTOPILOT", "category": "RUN_STATE", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "agent-state" ], "nextAction": "Review run state and resolve blocker if present.", "searchText": "care friction logs {\n \"last_run\": \"2026-04-03t09:00:00-06:00\",\n \"last_status\": \"blocked\",\n \"last_blocker\": \"discord webhook delivery issue (cannot read #dev channel); gws cli not found (cannot read care sheet)\",\n \"last_blocker_posted\": \"2026-04-03\",\n \"last_blocker_first_seen\": \"2026-04-03\"\n} ops/autopilot/inventory/agent-state/care-friction-logs.json https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/care-friction-logs.json autopilot run_state needs-review review run state and resolve blocker if present. autopilot review inventory agent-state" }, { "id": "doc:docs/designs/torus-product-architecture.md", "kind": "doc", "title": "CEO Plan: Torus Product Architecture", "summary": "---\nstatus: PROMOTED\n---\nGenerated by /plan-ceo-review on 2026-04-07", "path": "docs/designs/torus-product-architecture.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/designs/torus-product-architecture.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "designs" ], "searchText": "ceo plan: torus product architecture ---\nstatus: promoted\n---\ngenerated by /plan-ceo-review on 2026-04-07 docs/designs/torus-product-architecture.md https://github.com/torus-intelligence/torus/blob/staging/docs/designs/torus-product-architecture.md github architecture active docs architecture docs designs" }, { "id": "doc:docs/superpowers/plans/2026-05-28-change-order-redesign.md", "kind": "doc", "title": "Change-order redesign Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Make a change order a first-class workstream that lives inside the workspace shell as a tab, tells the customer-complaint → impact-graph → approvals story, and drops the `engineering_process_runs` middleman.\n**Architecture:** `agent_workstreams` owns the entire change-order substrate (story + impact graph FK + approvals FK). Backend functions (`runImpactAnalysis`, `draftEdits`, `mintApprovals`, `decideNode`, `createFromEmail`) take a `workstream_id`. New `backend/plugins/agent/change-orders/` module replaces the deleted `backend/plugins/processes/kinds/change_...", "path": "docs/superpowers/plans/2026-05-28-change-order-redesign.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-28-change-order-redesign.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "change-order redesign implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** make a change order a first-class workstream that lives inside the workspace shell as a tab, tells the customer-complaint → impact-graph → approvals story, and drops the `engineering_process_runs` middleman.\n**architecture:** `agent_workstreams` owns the entire change-order substrate (story + impact graph fk + approvals fk). backend functions (`runimpactanalysis`, `draftedits`, `mintapprovals`, `decidenode`, `createfromemail`) take a `workstream_id`. new `backend/plugins/agent/change-orders/` module replaces the deleted `backend/plugins/processes/kinds/change_... docs/superpowers/plans/2026-05-28-change-order-redesign.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-28-change-order-redesign.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/superpowers/specs/2026-05-28-change-order-redesign-design.md", "kind": "doc", "title": "Change-order redesign: tell the story inside the workspace", "summary": "**Date:** 2026-05-28\n**Status:** Draft, awaiting user review\n**Owner:** Rahul\nChange orders today don't tell a story and don't live in the workspace.", "path": "docs/superpowers/specs/2026-05-28-change-order-redesign-design.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-28-change-order-redesign-design.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "change-order redesign: tell the story inside the workspace **date:** 2026-05-28\n**status:** draft, awaiting user review\n**owner:** rahul\nchange orders today don't tell a story and don't live in the workspace. docs/superpowers/specs/2026-05-28-change-order-redesign-design.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-28-change-order-redesign-design.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:docs/superpowers/plans/2026-05-27-change-order-spine.md", "kind": "doc", "title": "Change-Order Spine Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Make EPC change-order propagation the product. One vendor change request → the system discovers the full impact graph, the agent drafts every cascading edit, each affected discipline lead reviews their node, and the request closes only when every node is decided. No more month-long Excel-tracked chase.\n**Architecture (three layers + one entry path):**\n- **Approvals plugin** (new) — Documenso-derived per-node signing primitive. Recipients with roles (signer / approver / cc / viewer), checklist-gated approve, sequential or parallel within a gate, full audit, HS2...", "path": "docs/superpowers/plans/2026-05-27-change-order-spine.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-27-change-order-spine.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "change-order spine implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** make epc change-order propagation the product. one vendor change request → the system discovers the full impact graph, the agent drafts every cascading edit, each affected discipline lead reviews their node, and the request closes only when every node is decided. no more month-long excel-tracked chase.\n**architecture (three layers + one entry path):**\n- **approvals plugin** (new) — documenso-derived per-node signing primitive. recipients with roles (signer / approver / cc / viewer), checklist-gated approve, sequential or parallel within a gate, full audit, hs2... docs/superpowers/plans/2026-05-27-change-order-spine.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-27-change-order-spine.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/architecture/cloud-tier-shadow-eval.md", "kind": "doc", "title": "Cloud LLM tier shadow eval", "summary": "W4-25 prep. **Not the flip** — cluster-C reverted an attempted standard-tier\nswap (gpt-5.4 → gpt-5.4-mini). This doc covers the eval gate that must pass\nbefore any future flip ships.\nThe actual run is a manual operator action. This doc describes the protocol;", "path": "docs/architecture/cloud-tier-shadow-eval.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/architecture/cloud-tier-shadow-eval.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "architecture" ], "searchText": "cloud llm tier shadow eval w4-25 prep. **not the flip** — cluster-c reverted an attempted standard-tier\nswap (gpt-5.4 → gpt-5.4-mini). this doc covers the eval gate that must pass\nbefore any future flip ships.\nthe actual run is a manual operator action. this doc describes the protocol; docs/architecture/cloud-tier-shadow-eval.md https://github.com/torus-intelligence/torus/blob/staging/docs/architecture/cloud-tier-shadow-eval.md github architecture active docs architecture docs architecture" }, { "id": "doc:ops/autopilot/docs/codex-runtime.md", "kind": "doc", "title": "Codex Automations -- Runtime Primer", "summary": "How Torus's autopilot runs on OpenAI Codex Automations.\nCurrent operating model: use Codex app Automations and have each scheduled\nprompt run the checked-in runner through `railway run`. See\n`ops/autopilot/docs/codex-automations.md` for the bring-up checklist and prompt", "path": "ops/autopilot/docs/codex-runtime.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/codex-runtime.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "codex automations -- runtime primer how torus's autopilot runs on openai codex automations.\ncurrent operating model: use codex app automations and have each scheduled\nprompt run the checked-in runner through `railway run`. see\n`ops/autopilot/docs/codex-automations.md` for the bring-up checklist and prompt ops/autopilot/docs/codex-runtime.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/codex-runtime.md autopilot autopilot active docs autopilot docs" }, { "id": "doc:ops/autopilot/docs/codex-automations.md", "kind": "doc", "title": "Codex Automations Bring-Up", "summary": "Current setup target: Codex app Automations running the autopilot runners on a\nschedule, with Railway supplying runtime environment variables.\nOpenAI's current Codex automation surface is the Codex app. Automations run\nrecurring tasks on a schedule and return results for review. OpenAI's own", "path": "ops/autopilot/docs/codex-automations.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/codex-automations.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "codex automations bring-up current setup target: codex app automations running the autopilot runners on a\nschedule, with railway supplying runtime environment variables.\nopenai's current codex automation surface is the codex app. automations run\nrecurring tasks on a schedule and return results for review. openai's own ops/autopilot/docs/codex-automations.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/codex-automations.md autopilot autopilot active docs autopilot docs" }, { "id": "doc:ops/autopilot/codex-environments/agent-output-analyzer.md", "kind": "doc", "title": "Codex Environment: agent-output-analyzer", "summary": "`0 15 * * 5` (Friday 3pm MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `DISCORD_LEADERSHIP_WEBHOOK_URL`", "path": "ops/autopilot/codex-environments/agent-output-analyzer.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/agent-output-analyzer.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: agent-output-analyzer `0 15 * * 5` (friday 3pm mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `discord_leadership_webhook_url` ops/autopilot/codex-environments/agent-output-analyzer.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/agent-output-analyzer.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/discord-mentions.md", "kind": "doc", "title": "Codex Environment: discord-mentions", "summary": "`*/10 * * * *`\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `DISCORD_BOT_TOKEN`", "path": "ops/autopilot/codex-environments/discord-mentions.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/discord-mentions.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: discord-mentions `*/10 * * * *`\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `discord_bot_token` ops/autopilot/codex-environments/discord-mentions.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/discord-mentions.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/engineering-deep-dive.md", "kind": "doc", "title": "Codex Environment: engineering-deep-dive", "summary": "`0 8 * * 6` (Saturday 8am MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `OPENAI_API_KEY` (for sub-agent LLM calls when wired)", "path": "ops/autopilot/codex-environments/engineering-deep-dive.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/engineering-deep-dive.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: engineering-deep-dive `0 8 * * 6` (saturday 8am mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `openai_api_key` (for sub-agent llm calls when wired) ops/autopilot/codex-environments/engineering-deep-dive.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/engineering-deep-dive.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/eod-brief.md", "kind": "doc", "title": "Codex Environment: eod-brief", "summary": "`0 18 * * *` (6pm MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `DISCORD_BOT_TOKEN`", "path": "ops/autopilot/codex-environments/eod-brief.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/eod-brief.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: eod-brief `0 18 * * *` (6pm mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `discord_bot_token` ops/autopilot/codex-environments/eod-brief.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/eod-brief.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/eow-review.md", "kind": "doc", "title": "Codex Environment: eow-review", "summary": "`0 16 * * 5` (Friday 4pm MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL` (#autopilot)\n- `DISCORD_LEADERSHIP_WEBHOOK_URL` (#leadership)", "path": "ops/autopilot/codex-environments/eow-review.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/eow-review.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: eow-review `0 16 * * 5` (friday 4pm mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url` (#autopilot)\n- `discord_leadership_webhook_url` (#leadership) ops/autopilot/codex-environments/eow-review.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/eow-review.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/feedback-triage.md", "kind": "doc", "title": "Codex Environment: feedback-triage", "summary": "`0 11 * * *` (daily 11am MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `DISCORD_BOT_TOKEN`", "path": "ops/autopilot/codex-environments/feedback-triage.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/feedback-triage.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: feedback-triage `0 11 * * *` (daily 11am mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `discord_bot_token` ops/autopilot/codex-environments/feedback-triage.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/feedback-triage.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/inbox-triage.md", "kind": "doc", "title": "Codex Environment: inbox-triage", "summary": "`0 7,13 * * *` (7am and 1pm MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `GMAIL_OAUTH_REFRESH_TOKEN`, `GMAIL_OAUTH_CLIENT_ID`, `GMAIL_OAUTH_CLIENT_SECRET`", "path": "ops/autopilot/codex-environments/inbox-triage.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/inbox-triage.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: inbox-triage `0 7,13 * * *` (7am and 1pm mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `gmail_oauth_refresh_token`, `gmail_oauth_client_id`, `gmail_oauth_client_secret` ops/autopilot/codex-environments/inbox-triage.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/inbox-triage.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/industry-event-response.md", "kind": "doc", "title": "Codex Environment: industry-event-response", "summary": "`*/30 13-23 * * 1-5` (every 30 min, business hours weekdays)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `DISCORD_LEADERSHIP_WEBHOOK_URL`", "path": "ops/autopilot/codex-environments/industry-event-response.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/industry-event-response.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: industry-event-response `*/30 13-23 * * 1-5` (every 30 min, business hours weekdays)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `discord_leadership_webhook_url` ops/autopilot/codex-environments/industry-event-response.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/industry-event-response.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/linear-first-pass.md", "kind": "doc", "title": "Codex Environment: linear-first-pass", "summary": "`*/5 * * * *` (every 5 minutes)\n`Torus-Intelligence/torus` -- branch `main`\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `LINEAR_API_KEY`", "path": "ops/autopilot/codex-environments/linear-first-pass.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/linear-first-pass.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: linear-first-pass `*/5 * * * *` (every 5 minutes)\n`torus-intelligence/torus` -- branch `main`\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `linear_api_key` ops/autopilot/codex-environments/linear-first-pass.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/linear-first-pass.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/marketing-deep-dive.md", "kind": "doc", "title": "Codex Environment: marketing-deep-dive", "summary": "`0 9 * * 6` (Saturday 9am MT)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `OPENAI_API_KEY` (sub-agent LLM calls)", "path": "ops/autopilot/codex-environments/marketing-deep-dive.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/marketing-deep-dive.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: marketing-deep-dive `0 9 * * 6` (saturday 9am mt)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `openai_api_key` (sub-agent llm calls) ops/autopilot/codex-environments/marketing-deep-dive.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/marketing-deep-dive.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "doc:ops/autopilot/codex-environments/standards-watch.md", "kind": "doc", "title": "Codex Environment: standards-watch", "summary": "`0 * * * *` (hourly)\n- `TZ=America/Denver` -- so MT labels in Discord posts and cron-firing time line up with the schedule comments\n- `DISCORD_WEBHOOK_URL`\n- `DATABASE_URL`", "path": "ops/autopilot/codex-environments/standards-watch.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/codex-environments/standards-watch.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "codex-environments" ], "searchText": "codex environment: standards-watch `0 * * * *` (hourly)\n- `tz=america/denver` -- so mt labels in discord posts and cron-firing time line up with the schedule comments\n- `discord_webhook_url`\n- `database_url` ops/autopilot/codex-environments/standards-watch.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/codex-environments/standards-watch.md autopilot autopilot active docs autopilot codex-environments" }, { "id": "skill:cold-outreach", "kind": "autopilot-skill", "title": "Cold Outreach", "summary": "---\nname: cold-outreach\ndescription: |\nDrafts first-touch outreach to EPC firm contacts. Apollo MCP / CSV.", "path": "ops/autopilot/skills/cold-outreach/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/cold-outreach/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "cold-outreach" ], "searchText": "cold outreach ---\nname: cold-outreach\ndescription: |\ndrafts first-touch outreach to epc firm contacts. apollo mcp / csv. ops/autopilot/skills/cold-outreach/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/cold-outreach/skill.md autopilot active autopilot skill skills cold-outreach" }, { "id": "doc:ops/autopilot/context/competitive.md", "kind": "doc", "title": "Competitive Landscape", "summary": "Source: web research + product analysis. Last synced: 2026-05-15.\nTorus is engineering intelligence for EPC/AEC. We read engineering documents (PFDs, P&IDs, equipment lists, specs, drawings, calc sheets, FEL2 studies) and codes (ASME, API, NFPA, CFIHOS V2.0, OSHA PSM), extract structured parameters, and run 50+ cross-discipline consistency checks. We generate deliverables (FEL2 reports, DOCX, diagrams) and learn firm-specific patterns from email and meeting observation.\nDifferentiators: domain primitives (CFIHOS V2.0 + OSHA PSM seeded, 478 ISA P&ID symbols, parameter extraction via GLiNER2 + GLiDRE), cross-discipline rule engine with L/G/C maturity, agent layer with 11 tools, trust filter for generated content.\nTorus does NOT do: jobsite vision/capture, contract review, process simulation...", "path": "ops/autopilot/context/competitive.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/competitive.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "competitive landscape source: web research + product analysis. last synced: 2026-05-15.\ntorus is engineering intelligence for epc/aec. we read engineering documents (pfds, p&ids, equipment lists, specs, drawings, calc sheets, fel2 studies) and codes (asme, api, nfpa, cfihos v2.0, osha psm), extract structured parameters, and run 50+ cross-discipline consistency checks. we generate deliverables (fel2 reports, docx, diagrams) and learn firm-specific patterns from email and meeting observation.\ndifferentiators: domain primitives (cfihos v2.0 + osha psm seeded, 478 isa p&id symbols, parameter extraction via gliner2 + glidre), cross-discipline rule engine with l/g/c maturity, agent layer with 11 tools, trust filter for generated content.\ntorus does not do: jobsite vision/capture, contract review, process simulation... ops/autopilot/context/competitive.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/competitive.md autopilot autopilot active docs autopilot context" }, { "id": "skill:competitor-comparison", "kind": "autopilot-skill", "title": "Competitor Comparison", "summary": "---\nname: competitor-comparison\ndescription: |\nUse to generate a one-page HTML comparison of Torus vs incumbent EPC tooling.", "path": "ops/autopilot/skills/competitor-comparison/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/competitor-comparison/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "competitor-comparison" ], "searchText": "competitor comparison ---\nname: competitor-comparison\ndescription: |\nuse to generate a one-page html comparison of torus vs incumbent epc tooling. ops/autopilot/skills/competitor-comparison/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/competitor-comparison/skill.md autopilot active autopilot skill skills competitor-comparison" }, { "id": "module:crm", "kind": "module", "title": "CRM", "summary": "Twenty-backed Companies, People, Opportunities, Tasks, Notes, warm accounts, follow-ups, and reviewed agent artifacts.", "href": "/crm", "source": "CRM", "status": "active", "tags": [ "crm", "twenty", "accounts", "follow-ups" ], "nextAction": "Open CRM or sync docs/autopilot artifacts into Twenty.", "searchText": "crm twenty-backed companies, people, opportunities, tasks, notes, warm accounts, follow-ups, and reviewed agent artifacts. /crm crm active open crm or sync docs/autopilot artifacts into twenty. crm twenty accounts follow-ups" }, { "id": "inventory:ops/autopilot/inventory/data-sources.md", "kind": "inventory", "title": "Data Sources", "summary": "> Updated weekly by `eng-weekend-deep-dive/osint-discovery`. Do not edit manually.\n**Last updated:** never\n**Updated by:** --\n| Source | Type | URL/Endpoint | Freshness | Cost | Integrated | Notes |", "path": "ops/autopilot/inventory/data-sources.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/data-sources.md", "source": "AUTOPILOT", "status": "active", "tags": [ "inventory", "inventory" ], "searchText": "data sources > updated weekly by `eng-weekend-deep-dive/osint-discovery`. do not edit manually.\n**last updated:** never\n**updated by:** --\n| source | type | url/endpoint | freshness | cost | integrated | notes | ops/autopilot/inventory/data-sources.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/data-sources.md autopilot active inventory inventory" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/schema-regen-pending.md", "kind": "doc", "title": "db/schema.sql Regeneration Pending — 2026-05-21 Weekly Audit", "summary": "Migrations 0105 through 0117 landed via `SKIP_DB_SCHEMA=1` (the pre-commit\nhook's documented escape hatch) because the worktree this audit ran in had\nno Railway credentials. The migration files themselves are committed and\nidempotent; `db/schema.sql` does NOT reflect them yet.", "path": "docs/superpowers/audits/2026-05-21-weekly/schema-regen-pending.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/schema-regen-pending.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "db/schema.sql regeneration pending — 2026-05-21 weekly audit migrations 0105 through 0117 landed via `skip_db_schema=1` (the pre-commit\nhook's documented escape hatch) because the worktree this audit ran in had\nno railway credentials. the migration files themselves are committed and\nidempotent; `db/schema.sql` does not reflect them yet. docs/superpowers/audits/2026-05-21-weekly/schema-regen-pending.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/schema-regen-pending.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/deployment.md", "kind": "doc", "title": "Deployment", "summary": "Two Railway environments: **production** and **staging**. Each has its own database, domains, and deployments.\n| Service | Root Dir | Builder | Domain | Branch |\n|---------|----------|---------|--------|--------|\n| t-backend | repo root | Dockerfile.backend | api.usetorus.com | main |", "path": "docs/deployment.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/deployment.md", "source": "GITHUB", "category": "DEPLOYMENT", "status": "active", "tags": [ "docs", "deployment", "docs" ], "searchText": "deployment two railway environments: **production** and **staging**. each has its own database, domains, and deployments.\n| service | root dir | builder | domain | branch |\n|---------|----------|---------|--------|--------|\n| t-backend | repo root | dockerfile.backend | api.usetorus.com | main | docs/deployment.md https://github.com/torus-intelligence/torus/blob/staging/docs/deployment.md github deployment active docs deployment docs" }, { "id": "inventory:ops/autopilot/inventory/deps.json", "kind": "inventory", "title": "Deps", "summary": "{\n \"_meta\": {\n \"description\": \"Living dependency graph across all Torus repos. Updated weekly by eng-weekend-deep-dive/research-review.\",\n \"last_updated\": null,\n \"updated_by\": null\n },\n \"repos\": {}\n}", "path": "ops/autopilot/inventory/deps.json", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/deps.json", "source": "AUTOPILOT", "status": "active", "tags": [ "inventory", "inventory" ], "searchText": "deps {\n \"_meta\": {\n \"description\": \"living dependency graph across all torus repos. updated weekly by eng-weekend-deep-dive/research-review.\",\n \"last_updated\": null,\n \"updated_by\": null\n },\n \"repos\": {}\n} ops/autopilot/inventory/deps.json https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/deps.json autopilot active inventory inventory" }, { "id": "doc:ops/autopilot/docs/det-vs-judgment.md", "kind": "doc", "title": "Deterministic vs Judgment Steps (Autopilot)", "summary": "Phase 7d principle adapted from gbrain's \"Minions vs sub-agents\" routing\nrule. Mechanical work does NOT go through an LLM.\n- **Deterministic** (`[det]`): same input produces same output via pure\ncode. File reads, API calls with known arguments, JSON writes,", "path": "ops/autopilot/docs/det-vs-judgment.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/det-vs-judgment.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "deterministic vs judgment steps (autopilot) phase 7d principle adapted from gbrain's \"minions vs sub-agents\" routing\nrule. mechanical work does not go through an llm.\n- **deterministic** (`[det]`): same input produces same output via pure\ncode. file reads, api calls with known arguments, json writes, ops/autopilot/docs/det-vs-judgment.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/det-vs-judgment.md autopilot autopilot active docs autopilot docs" }, { "id": "doc:docs/superpowers/specs/2026-05-20-disaster-recovery-warm-standby-design.md", "kind": "doc", "title": "Disaster Recovery -- Warm Standby on Fly.io", "summary": "Single-switch failover from Railway (primary) to Fly.io (warm standby) when Railway -- or the colocation/cloud underneath it -- goes down. Triggered by the May 2026 GCP-side incident that took Railway offline mid-workday.\nThe entire production stack runs on Railway: backend (Bun), frontend (SvelteKit), sidecar (Python ML), Postgres (with pgvector + pg_trgm), and S3-compatible object storage. Railway operates its own colocation in four regions but has GCP failover paths, so a major GCP incident can still ripple into Railway availability. When that happens, every client loses access -- read and write -- with no recourse beyond \"wait for Railway.\"\nWe want a warm standby that an on-call engineer can promote in under five minutes with a single command, with clear acceptance that some seconds o...", "path": "docs/superpowers/specs/2026-05-20-disaster-recovery-warm-standby-design.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-20-disaster-recovery-warm-standby-design.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "disaster recovery -- warm standby on fly.io single-switch failover from railway (primary) to fly.io (warm standby) when railway -- or the colocation/cloud underneath it -- goes down. triggered by the may 2026 gcp-side incident that took railway offline mid-workday.\nthe entire production stack runs on railway: backend (bun), frontend (sveltekit), sidecar (python ml), postgres (with pgvector + pg_trgm), and s3-compatible object storage. railway operates its own colocation in four regions but has gcp failover paths, so a major gcp incident can still ripple into railway availability. when that happens, every client loses access -- read and write -- with no recourse beyond \"wait for railway.\"\nwe want a warm standby that an on-call engineer can promote in under five minutes with a single command, with clear acceptance that some seconds o... docs/superpowers/specs/2026-05-20-disaster-recovery-warm-standby-design.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-20-disaster-recovery-warm-standby-design.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:ops/autopilot/context/channels.md", "kind": "doc", "title": "Discord Channels", "summary": "| Channel | Purpose | Agents That Use It |\n|---------|---------|-------------------|\n| #dev | Dev chat + bug reports | CARE, EOW Review |\n| #dailies | Daily standups from everyone | EOW Review |", "path": "ops/autopilot/context/channels.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/channels.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "discord channels | channel | purpose | agents that use it |\n|---------|---------|-------------------|\n| #dev | dev chat + bug reports | care, eow review |\n| #dailies | daily standups from everyone | eow review | ops/autopilot/context/channels.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/channels.md autopilot autopilot active docs autopilot context" }, { "id": "runner:discord-mentions", "kind": "autopilot-runner", "title": "Discord Mentions", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/discord-mentions/run.ts.", "path": "ops/autopilot/runners/discord-mentions/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/discord-mentions/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "discord-mentions" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "discord mentions scheduled or operator-run autopilot workflow at ops/autopilot/runners/discord-mentions/run.ts. ops/autopilot/runners/discord-mentions/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/discord-mentions/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner discord-mentions" }, { "id": "skill:discord-mentions", "kind": "autopilot-skill", "title": "Discord Mentions", "summary": "---\nname: discord-mentions\ndescription: On-demand (10-min poll). Listens for @-mentions in a configured Discord channel; executes commands; reaction-emoji gate for state-changing ops.\n---", "path": "ops/autopilot/skills/discord-mentions/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/discord-mentions/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "discord-mentions" ], "searchText": "discord mentions ---\nname: discord-mentions\ndescription: on-demand (10-min poll). listens for @-mentions in a configured discord channel; executes commands; reaction-emoji gate for state-changing ops.\n--- ops/autopilot/skills/discord-mentions/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/discord-mentions/skill.md autopilot active autopilot skill skills discord-mentions" }, { "id": "module:docs", "kind": "module", "title": "Docs / Runbooks", "summary": "GitHub-backed operational docs, runbooks, architecture notes, generated reports, and Autopilot context.", "href": "#docs", "source": "GITHUB", "status": "active", "tags": [ "docs", "runbooks", "github" ], "searchText": "docs / runbooks github-backed operational docs, runbooks, architecture notes, generated reports, and autopilot context. #docs github active docs runbooks github" }, { "id": "doc:docs/inspirations/documenso.md", "kind": "doc", "title": "Documenso (documenso/documenso)", "summary": "**Source:** https://github.com/documenso/documenso -- https://documenso.com\n**License:** AGPL-3.0 (read for inspiration only; no code copy)\n**First documented:** 2026-05-27\n**Last reviewed:** 2026-05-27 (v2.11.0, released 2026-05-13)", "path": "docs/inspirations/documenso.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/inspirations/documenso.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "inspirations" ], "searchText": "documenso (documenso/documenso) **source:** https://github.com/documenso/documenso -- https://documenso.com\n**license:** agpl-3.0 (read for inspiration only; no code copy)\n**first documented:** 2026-05-27\n**last reviewed:** 2026-05-27 (v2.11.0, released 2026-05-13) docs/inspirations/documenso.md https://github.com/torus-intelligence/torus/blob/staging/docs/inspirations/documenso.md github runbook active docs runbook docs inspirations" }, { "id": "skill:email-handler", "kind": "autopilot-skill", "title": "Email Handler", "summary": "---\nname: email-handler\ndescription: |\nUse when processing $AUTOPILOT_USER's Gmail inbox -- triaging emails, drafting replies,", "path": "ops/autopilot/skills/email-handler/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/email-handler/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "email-handler" ], "searchText": "email handler ---\nname: email-handler\ndescription: |\nuse when processing $autopilot_user's gmail inbox -- triaging emails, drafting replies, ops/autopilot/skills/email-handler/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/email-handler/skill.md autopilot active autopilot skill skills email-handler" }, { "id": "doc:docs/email-pipeline.md", "kind": "doc", "title": "Email Pipeline", "summary": "1. **Parse address** -- `projectcedar@bv.usetorus.cc` → firm \"bv\", project \"projectcedar\". Auto-creates firm/project if first seen.\n2. **Spam pre-filter** -- Reject auto-replies (X-Auto-Reply), OOB (subject match), newsletters (List-Unsubscribe). Stored in `filtered_noise` collection.\n3. **Fetch full email** -- Resend webhook is just a notification. Fetches body + attachments via Resend API.\n4. **Classify** -- Classification router (`router.ts`): deterministic rules first (subject patterns, attachment types), nano LLM fallback for ambiguous cases. Dual classification: process type (RFI, submittal, procurement, etc.) + document type per attachment (100+ types). Low confidence (<0.6) flagged for PM review.", "path": "docs/email-pipeline.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/email-pipeline.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "email pipeline 1. **parse address** -- `projectcedar@bv.usetorus.cc` → firm \"bv\", project \"projectcedar\". auto-creates firm/project if first seen.\n2. **spam pre-filter** -- reject auto-replies (x-auto-reply), oob (subject match), newsletters (list-unsubscribe). stored in `filtered_noise` collection.\n3. **fetch full email** -- resend webhook is just a notification. fetches body + attachments via resend api.\n4. **classify** -- classification router (`router.ts`): deterministic rules first (subject patterns, attachment types), nano llm fallback for ambiguous cases. dual classification: process type (rfi, submittal, procurement, etc.) + document type per attachment (100+ types). low confidence (<0.6) flagged for pm review. docs/email-pipeline.md https://github.com/torus-intelligence/torus/blob/staging/docs/email-pipeline.md github runbook active docs runbook docs" }, { "id": "skill:eow-review", "kind": "autopilot-skill", "title": "End-of-Week Review", "summary": "---\nname: eow-review\ndescription: Friday afternoon dual-channel weekly digest. Public wins to #autopilot; private accountability to #leadership.\n---", "path": "ops/autopilot/skills/eow-review/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/eow-review/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "eow-review" ], "searchText": "end-of-week review ---\nname: eow-review\ndescription: friday afternoon dual-channel weekly digest. public wins to #autopilot; private accountability to #leadership.\n--- ops/autopilot/skills/eow-review/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/eow-review/skill.md autopilot active autopilot skill skills eow-review" }, { "id": "runner:engineering-deep-dive", "kind": "autopilot-runner", "title": "Engineering Deep Dive", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/engineering-deep-dive/run.ts.", "path": "ops/autopilot/runners/engineering-deep-dive/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/engineering-deep-dive/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "engineering-deep-dive" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "engineering deep dive scheduled or operator-run autopilot workflow at ops/autopilot/runners/engineering-deep-dive/run.ts. ops/autopilot/runners/engineering-deep-dive/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/engineering-deep-dive/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner engineering-deep-dive" }, { "id": "skill:engineering-deep-dive", "kind": "autopilot-skill", "title": "Engineering Deep Dive", "summary": "---\nname: engineering-deep-dive\ndescription: Saturday weekly engineering review. Orchestrates 7 sub-agents. Continue-on-failure with per-sub-agent time budget.\n---", "path": "ops/autopilot/skills/engineering-deep-dive/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/engineering-deep-dive/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "engineering-deep-dive" ], "searchText": "engineering deep dive ---\nname: engineering-deep-dive\ndescription: saturday weekly engineering review. orchestrates 7 sub-agents. continue-on-failure with per-sub-agent time budget.\n--- ops/autopilot/skills/engineering-deep-dive/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/engineering-deep-dive/skill.md autopilot active autopilot skill skills engineering-deep-dive" }, { "id": "doc:docs/engineering-intelligence.md", "kind": "doc", "title": "Engineering Intelligence Engine", "summary": "Reads engineering documents, extracts every identifiable parameter, maps relationships across disciplines, builds a knowledge graph, and runs deterministic consistency checks. Finds the inconsistency between a process datasheet and a piping spec before it becomes a rework order.\nUnion views over existing tables + new graph tables. Everything is connected.\n**Node types:** parameter, document, equipment, system, discipline, person_org, material, code_standard, goal, flag, rfi, process\n**Edge types:** extracted_from, belongs_to, part_of, produced_by, owned_by, constrains, references, feeds_into, governed_by, supplied_by, serves, requires, flagged_by, triggered_rule, resolved_by, assigned_to, has_role, prerequisite_for", "path": "docs/engineering-intelligence.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/engineering-intelligence.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "engineering intelligence engine reads engineering documents, extracts every identifiable parameter, maps relationships across disciplines, builds a knowledge graph, and runs deterministic consistency checks. finds the inconsistency between a process datasheet and a piping spec before it becomes a rework order.\nunion views over existing tables + new graph tables. everything is connected.\n**node types:** parameter, document, equipment, system, discipline, person_org, material, code_standard, goal, flag, rfi, process\n**edge types:** extracted_from, belongs_to, part_of, produced_by, owned_by, constrains, references, feeds_into, governed_by, supplied_by, serves, requires, flagged_by, triggered_rule, resolved_by, assigned_to, has_role, prerequisite_for docs/engineering-intelligence.md https://github.com/torus-intelligence/torus/blob/staging/docs/engineering-intelligence.md github runbook active docs runbook docs" }, { "id": "runner:eod-brief", "kind": "autopilot-runner", "title": "Eod Brief", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/eod-brief/run.ts.", "path": "ops/autopilot/runners/eod-brief/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/eod-brief/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "eod-brief" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "eod brief scheduled or operator-run autopilot workflow at ops/autopilot/runners/eod-brief/run.ts. ops/autopilot/runners/eod-brief/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/eod-brief/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner eod-brief" }, { "id": "skill:eod-brief", "kind": "autopilot-skill", "title": "EOD Brief", "summary": "---\nname: eod-brief\ndescription: Once-a-day end-of-day digest. Reads Linear + Gmail + Discord; posts to #autopilot and DMs Rahul a 1-liner.\n---", "path": "ops/autopilot/skills/eod-brief/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/eod-brief/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "eod-brief" ], "searchText": "eod brief ---\nname: eod-brief\ndescription: once-a-day end-of-day digest. reads linear + gmail + discord; posts to #autopilot and dms rahul a 1-liner.\n--- ops/autopilot/skills/eod-brief/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/eod-brief/skill.md autopilot active autopilot skill skills eod-brief" }, { "id": "runner:eod-sweep", "kind": "autopilot-runner", "title": "Eod Sweep", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/eod-sweep/run.ts.", "path": "ops/autopilot/runners/eod-sweep/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/eod-sweep/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "eod-sweep" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "eod sweep scheduled or operator-run autopilot workflow at ops/autopilot/runners/eod-sweep/run.ts. ops/autopilot/runners/eod-sweep/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/eod-sweep/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner eod-sweep" }, { "id": "skill:eod-sweep", "kind": "autopilot-skill", "title": "EOD Sweep", "summary": "---\nname: eod-sweep\ndescription: |\nEnd-of-day orchestrator. Reviews the day's commits, product changelog, ML metrics, agent state files, and emits a Discord digest with BLOCKED / RAN TODAY / SILENT sections.", "path": "ops/autopilot/skills/eod-sweep/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/eod-sweep/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "eod-sweep" ], "searchText": "eod sweep ---\nname: eod-sweep\ndescription: |\nend-of-day orchestrator. reviews the day's commits, product changelog, ml metrics, agent state files, and emits a discord digest with blocked / ran today / silent sections. ops/autopilot/skills/eod-sweep/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/eod-sweep/skill.md autopilot active autopilot skill skills eod-sweep" }, { "id": "runner:eow-review", "kind": "autopilot-runner", "title": "Eow Review", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/eow-review/run.ts.", "path": "ops/autopilot/runners/eow-review/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/eow-review/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "eow-review" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "eow review scheduled or operator-run autopilot workflow at ops/autopilot/runners/eow-review/run.ts. ops/autopilot/runners/eow-review/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/eow-review/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner eow-review" }, { "id": "skill:epc-sourcing", "kind": "autopilot-skill", "title": "EPC Sourcing", "summary": "---\nname: epc-sourcing\ndescription: Use when sourcing EPC (Engineering, Procurement, Construction) and engineering firms in CA or CO from public databases. Outputs linked CSVs of firms, projects, and people.\n---", "path": "ops/autopilot/skills/epc-sourcing/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/epc-sourcing/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "epc-sourcing" ], "searchText": "epc sourcing ---\nname: epc-sourcing\ndescription: use when sourcing epc (engineering, procurement, construction) and engineering firms in ca or co from public databases. outputs linked csvs of firms, projects, and people.\n--- ops/autopilot/skills/epc-sourcing/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/epc-sourcing/skill.md autopilot active autopilot skill skills epc-sourcing" }, { "id": "skill:event-materials", "kind": "autopilot-skill", "title": "Event Materials", "summary": "---\nname: event-materials\ndescription: |\nOn-demand. One message in, one HTML out, no clarifying questions. Print-ready A4/Letter one-pager generator.", "path": "ops/autopilot/skills/event-materials/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/event-materials/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "event-materials" ], "searchText": "event materials ---\nname: event-materials\ndescription: |\non-demand. one message in, one html out, no clarifying questions. print-ready a4/letter one-pager generator. ops/autopilot/skills/event-materials/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/event-materials/skill.md autopilot active autopilot skill skills event-materials" }, { "id": "doc:ops/autopilot/context/data-sources.md", "kind": "doc", "title": "External Data Sources -- Reference", "summary": "Static catalog of every external API, RSS feed, scraper, and database autopilot agents touch. Runtime registry of *discovered* sources lives at `ops/autopilot/inventory/data-sources.md` (managed by `engineering-deep-dive/standards-watch`).\n| Source | Auth | Endpoint | Used by | Rate limit |\n|--------|------|----------|---------|------------|\n| Linear | API key (`Authorization: `) | `api.linear.app/graphql` | linear-first-pass, email-handler, feedback-triage, discord-mentions, eod-brief | 1500/h |", "path": "ops/autopilot/context/data-sources.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/data-sources.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "external data sources -- reference static catalog of every external api, rss feed, scraper, and database autopilot agents touch. runtime registry of *discovered* sources lives at `ops/autopilot/inventory/data-sources.md` (managed by `engineering-deep-dive/standards-watch`).\n| source | auth | endpoint | used by | rate limit |\n|--------|------|----------|---------|------------|\n| linear | api key (`authorization: `) | `api.linear.app/graphql` | linear-first-pass, email-handler, feedback-triage, discord-mentions, eod-brief | 1500/h | ops/autopilot/context/data-sources.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/data-sources.md autopilot autopilot active docs autopilot context" }, { "id": "doc:docs/extraction-pipeline.md", "kind": "doc", "title": "Extraction Pipeline", "summary": "Entry point: `runExtractionPipeline()` in `backend/plugins/engineering/extraction-pipeline.ts`.\nTriggered automatically on document upload (via `dataroom/pipeline.ts`). Can also be triggered manually via POST `/plugins/engineering/ingest`.\n`tier-classifier.ts` routes documents by format:\n| Tier | Format | Strategy |", "path": "docs/extraction-pipeline.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/extraction-pipeline.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "extraction pipeline entry point: `runextractionpipeline()` in `backend/plugins/engineering/extraction-pipeline.ts`.\ntriggered automatically on document upload (via `dataroom/pipeline.ts`). can also be triggered manually via post `/plugins/engineering/ingest`.\n`tier-classifier.ts` routes documents by format:\n| tier | format | strategy | docs/extraction-pipeline.md https://github.com/torus-intelligence/torus/blob/staging/docs/extraction-pipeline.md github runbook active docs runbook docs" }, { "id": "runner:feedback-triage", "kind": "autopilot-runner", "title": "Feedback Triage", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/feedback-triage/run.ts.", "path": "ops/autopilot/runners/feedback-triage/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/feedback-triage/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "feedback-triage" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "feedback triage scheduled or operator-run autopilot workflow at ops/autopilot/runners/feedback-triage/run.ts. ops/autopilot/runners/feedback-triage/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/feedback-triage/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner feedback-triage" }, { "id": "skill:feedback-triage", "kind": "autopilot-skill", "title": "Feedback Triage", "summary": "---\nname: feedback-triage\ndescription: |\nDaily scan of Discord #dev for bugs and friction; classifies, dedupes against Linear, files novel tickets, closes the loop in-thread.", "path": "ops/autopilot/skills/feedback-triage/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/feedback-triage/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "feedback-triage" ], "searchText": "feedback triage ---\nname: feedback-triage\ndescription: |\ndaily scan of discord #dev for bugs and friction; classifies, dedupes against linear, files novel tickets, closes the loop in-thread. ops/autopilot/skills/feedback-triage/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/feedback-triage/skill.md autopilot active autopilot skill skills feedback-triage" }, { "id": "doc:ops/autopilot/context/figma.md", "kind": "doc", "title": "Figma Workspace", "summary": "\nplanKey: PENDING -- Figma MCP tools not available in current environment\nThe marketing-visuals skill requires these fonts loaded via `figma.loadFontAsync()` before setting any text:\n| Font | Weight | Source |", "path": "ops/autopilot/context/figma.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/figma.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "figma workspace \nplankey: pending -- figma mcp tools not available in current environment\nthe marketing-visuals skill requires these fonts loaded via `figma.loadfontasync()` before setting any text:\n| font | weight | source | ops/autopilot/context/figma.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/figma.md autopilot autopilot active docs autopilot context" }, { "id": "skill:firm-snapshot", "kind": "autopilot-skill", "title": "Firm Snapshot", "summary": "---\nname: firm-snapshot\ndescription: |\nUse to generate a one-page HTML briefing on a single EPC firm.", "path": "ops/autopilot/skills/firm-snapshot/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/firm-snapshot/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "firm-snapshot" ], "searchText": "firm snapshot ---\nname: firm-snapshot\ndescription: |\nuse to generate a one-page html briefing on a single epc firm. ops/autopilot/skills/firm-snapshot/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/firm-snapshot/skill.md autopilot active autopilot skill skills firm-snapshot" }, { "id": "skill:fit-scorer", "kind": "autopilot-skill", "title": "Fit Scorer", "summary": "---\nname: fit-scorer\ndescription: |\nUse to score an inbound RFP, RFI, or inquiry against Torus's ICP.", "path": "ops/autopilot/skills/fit-scorer/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/fit-scorer/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "fit-scorer" ], "searchText": "fit scorer ---\nname: fit-scorer\ndescription: |\nuse to score an inbound rfp, rfi, or inquiry against torus's icp. ops/autopilot/skills/fit-scorer/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/fit-scorer/skill.md autopilot active autopilot skill skills fit-scorer" }, { "id": "doc:docs/superpowers/plans/2026-05-21-fix-projects-sidebar-and-merge-sl.md", "kind": "doc", "title": "Fix Projects Sidebar (Staging) + Consolidate Duplicate S&L Firm", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Restore real projects (Bantam, Project Feather, FEL2 Sargent Lundy) in the admin sidebar on staging by sending identity headers from server loaders to the W4-3-hardened backend, then merge the duplicate Sargent & Lundy firm row so admin sees one canonical account.\n**Architecture:** The W4-3 sweep (PR #44) added `requireFirmAccess` to firm-scoped backend routes — they now require `X-Torus-Firm-Id` / `X-Torus-User-Email` / `X-Torus-User-Role` headers. The frontend SvelteKit server loaders still call `platformGet` from `$lib/api`, which only sends `Authorization:...", "path": "docs/superpowers/plans/2026-05-21-fix-projects-sidebar-and-merge-sl.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-fix-projects-sidebar-and-merge-sl.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "fix projects sidebar (staging) + consolidate duplicate s&l firm > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** restore real projects (bantam, project feather, fel2 sargent lundy) in the admin sidebar on staging by sending identity headers from server loaders to the w4-3-hardened backend, then merge the duplicate sargent & lundy firm row so admin sees one canonical account.\n**architecture:** the w4-3 sweep (pr #44) added `requirefirmaccess` to firm-scoped backend routes — they now require `x-torus-firm-id` / `x-torus-user-email` / `x-torus-user-role` headers. the frontend sveltekit server loaders still call `platformget` from `$lib/api`, which only sends `authorization:... docs/superpowers/plans/2026-05-21-fix-projects-sidebar-and-merge-sl.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-fix-projects-sidebar-and-merge-sl.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/superpowers/plans/2026-05-21-larger-cleanups.md", "kind": "doc", "title": "Four-Track Cleanup Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Close out four longstanding refactor items surfaced by the 2026-05-21 dead-code audit: signal-strength scoring, two large file splits, and multi-replica session backing.\n**Architecture:** Four independent tracks. Each ships on its own PR. No cross-track dependencies — pick any order. Tracks 1 and 4 are backend (Bun + Postgres). Tracks 2 and 3 are SvelteKit frontend.\n**Tech Stack:** Bun, TypeScript, Postgres, SvelteKit 5, postgres-js. Optional Redis (track 4 only, behind env flag).", "path": "docs/superpowers/plans/2026-05-21-larger-cleanups.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-larger-cleanups.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "four-track cleanup implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** close out four longstanding refactor items surfaced by the 2026-05-21 dead-code audit: signal-strength scoring, two large file splits, and multi-replica session backing.\n**architecture:** four independent tracks. each ships on its own pr. no cross-track dependencies — pick any order. tracks 1 and 4 are backend (bun + postgres). tracks 2 and 3 are sveltekit frontend.\n**tech stack:** bun, typescript, postgres, sveltekit 5, postgres-js. optional redis (track 4 only, behind env flag). docs/superpowers/plans/2026-05-21-larger-cleanups.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-larger-cleanups.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/frontend.md", "kind": "doc", "title": "Frontend (app.usetorus.com)", "summary": "SvelteKit 5 app with adapter-node. Deployed via Docker on Railway.\n- **Email + password:** Login form calls `/plugins/auth/login`; password verification lives in the backend auth plugin.\n- **Admin bypass:** `admin@usetorus.com` / `adminadmin` still exists for demo/admin access, but it must resolve a real firm UUID before a session is minted.\n- **Recent-login magic link fallback:** Failed password login can issue a one-time email link only for accounts with `last_login_at` inside the configured recent-login window.", "path": "docs/frontend.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/frontend.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "frontend (app.usetorus.com) sveltekit 5 app with adapter-node. deployed via docker on railway.\n- **email + password:** login form calls `/plugins/auth/login`; password verification lives in the backend auth plugin.\n- **admin bypass:** `admin@usetorus.com` / `adminadmin` still exists for demo/admin access, but it must resolve a real firm uuid before a session is minted.\n- **recent-login magic link fallback:** failed password login can issue a one-time email link only for accounts with `last_login_at` inside the configured recent-login window. docs/frontend.md https://github.com/torus-intelligence/torus/blob/staging/docs/frontend.md github runbook active docs runbook docs" }, { "id": "integration:github-docs", "kind": "integration", "title": "GitHub Docs Index", "summary": "GitHub remains the source of truth; Ops indexes summaries and links for humans, CRM, and agents.", "source": "GITHUB", "status": "active", "tags": [ "github", "docs", "search" ], "searchText": "github docs index github remains the source of truth; ops indexes summaries and links for humans, crm, and agents. github active github docs search" }, { "id": "doc:ops/autopilot/docs/gws-setup.md", "kind": "doc", "title": "Google Workspace CLI Setup", "summary": "`gws` is our CLI for Google Drive, Sheets, Docs, and Calendar. It replaces the Google Drive MCP connector.\nTorus uses GWS for Drive, Sheets, Docs, Calendar, and Gmail.\nVerify: `gws --version` should show 0.22.5+.\n1. Copy the client secret from this repo:", "path": "ops/autopilot/docs/gws-setup.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/gws-setup.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "google workspace cli setup `gws` is our cli for google drive, sheets, docs, and calendar. it replaces the google drive mcp connector.\ntorus uses gws for drive, sheets, docs, calendar, and gmail.\nverify: `gws --version` should show 0.22.5+.\n1. copy the client secret from this repo: ops/autopilot/docs/gws-setup.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/gws-setup.md autopilot autopilot active docs autopilot docs" }, { "id": "doc:docs/architecture/graduation-ledger.md", "kind": "doc", "title": "Graduation Ledger Architecture", "summary": "The graduation ledger is the cross-system audit log for phase transitions\n(`liquid` → `glass` → `crystal` → `demoted`). One row per transition, across\nevery substrate that uses the L/G/C phase model.\nBefore migration 0056, the ledger was a single table with polymorphic refs:", "path": "docs/architecture/graduation-ledger.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/architecture/graduation-ledger.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "architecture" ], "searchText": "graduation ledger architecture the graduation ledger is the cross-system audit log for phase transitions\n(`liquid` → `glass` → `crystal` → `demoted`). one row per transition, across\nevery substrate that uses the l/g/c phase model.\nbefore migration 0056, the ledger was a single table with polymorphic refs: docs/architecture/graduation-ledger.md https://github.com/torus-intelligence/torus/blob/staging/docs/architecture/graduation-ledger.md github architecture active docs architecture docs architecture" }, { "id": "integration:granola", "kind": "integration", "title": "Granola Meeting Intelligence", "summary": "Meeting notes should create attributed notes, follow-up tasks, pain/buying signals, and reviewable Ops Artifacts.", "source": "GRANOLA", "status": "planned", "tags": [ "granola", "meetings", "signals", "follow-ups" ], "nextAction": "Add Granola sync through the Ops API bridge, not direct browser automation.", "searchText": "granola meeting intelligence meeting notes should create attributed notes, follow-up tasks, pain/buying signals, and reviewable ops artifacts. granola planned add granola sync through the ops api bridge, not direct browser automation. granola meetings signals follow-ups" }, { "id": "doc:ops/autopilot/docs/hackathon-guide.md", "kind": "doc", "title": "Hackathon Guide", "summary": "You have your own skill directory, your own branch, and Claude to help you build. Here's everything you need to know.\n- **Your skill** lives in `skills/hackathon-/` -- scaffolded for you by skill-builder\n- **Your branch** is `hackathon/` -- all your work goes here\n- **Claude** writes the code -- you describe what you want the agent to do", "path": "ops/autopilot/docs/hackathon-guide.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/hackathon-guide.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "hackathon guide you have your own skill directory, your own branch, and claude to help you build. here's everything you need to know.\n- **your skill** lives in `skills/hackathon-/` -- scaffolded for you by skill-builder\n- **your branch** is `hackathon/` -- all your work goes here\n- **claude** writes the code -- you describe what you want the agent to do ops/autopilot/docs/hackathon-guide.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/hackathon-guide.md autopilot autopilot active docs autopilot docs" }, { "id": "autopilot-output:ops/autopilot/inventory/agent-state/hiring-autofilter.json", "kind": "autopilot-output", "title": "Hiring Autofilter", "summary": "{\n \"last_run\": \"2026-04-03T00:05:00Z\",\n \"last_status\": \"DONE_EMPTY\",\n \"last_blocker\": null,\n \"last_blocker_posted\": null,\n \"last_blocker_first_seen\": null\n}", "path": "ops/autopilot/inventory/agent-state/hiring-autofilter.json", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/hiring-autofilter.json", "source": "AUTOPILOT", "category": "RUN_STATE", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "agent-state" ], "nextAction": "Review run state and resolve blocker if present.", "searchText": "hiring autofilter {\n \"last_run\": \"2026-04-03t00:05:00z\",\n \"last_status\": \"done_empty\",\n \"last_blocker\": null,\n \"last_blocker_posted\": null,\n \"last_blocker_first_seen\": null\n} ops/autopilot/inventory/agent-state/hiring-autofilter.json https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/hiring-autofilter.json autopilot run_state needs-review review run state and resolve blocker if present. autopilot review inventory agent-state" }, { "id": "skill:hiring-autofilter", "kind": "autopilot-skill", "title": "Hiring Autofilter", "summary": "\n\n---\nname: hiring-autofilter", "path": "ops/autopilot/skills/hiring-autofilter/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/hiring-autofilter/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "hiring-autofilter" ], "searchText": "hiring autofilter \n\n---\nname: hiring-autofilter ops/autopilot/skills/hiring-autofilter/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/hiring-autofilter/skill.md autopilot active autopilot skill skills hiring-autofilter" }, { "id": "doc:ops/autopilot/context/icp.md", "kind": "doc", "title": "Ideal Customer Profile (Torus)", "summary": "- EPC firms in process / petchem / power, $50M+ project scope\n- FEL2 or FEL3 phase\n- Strict code referencing (ASME B31.x, API 650/620, NFPA 70/72, NEC)\n- Engineering manager / project engineer / discipline lead is the buyer", "path": "ops/autopilot/context/icp.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/icp.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "ideal customer profile (torus) - epc firms in process / petchem / power, $50m+ project scope\n- fel2 or fel3 phase\n- strict code referencing (asme b31.x, api 650/620, nfpa 70/72, nec)\n- engineering manager / project engineer / discipline lead is the buyer ops/autopilot/context/icp.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/icp.md autopilot autopilot active docs autopilot context" }, { "id": "runner:inbox-triage", "kind": "autopilot-runner", "title": "Inbox Triage", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/inbox-triage/run.ts.", "path": "ops/autopilot/runners/inbox-triage/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/inbox-triage/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "inbox-triage" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "inbox triage scheduled or operator-run autopilot workflow at ops/autopilot/runners/inbox-triage/run.ts. ops/autopilot/runners/inbox-triage/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/inbox-triage/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner inbox-triage" }, { "id": "runner:industry-event-response", "kind": "autopilot-runner", "title": "Industry Event Response", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/industry-event-response/run.ts.", "path": "ops/autopilot/runners/industry-event-response/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/industry-event-response/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "industry-event-response" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "industry event response scheduled or operator-run autopilot workflow at ops/autopilot/runners/industry-event-response/run.ts. ops/autopilot/runners/industry-event-response/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/industry-event-response/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner industry-event-response" }, { "id": "skill:industry-event-response", "kind": "autopilot-skill", "title": "Industry Event Response", "summary": "---\nname: industry-event-response\ndescription: |\nWhen an industry event hits, find affected projects + contacts, draft outreach in three flavors, queue for send.", "path": "ops/autopilot/skills/industry-event-response/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/industry-event-response/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "industry-event-response" ], "searchText": "industry event response ---\nname: industry-event-response\ndescription: |\nwhen an industry event hits, find affected projects + contacts, draft outreach in three flavors, queue for send. ops/autopilot/skills/industry-event-response/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/industry-event-response/skill.md autopilot active autopilot skill skills industry-event-response" }, { "id": "doc:ops/autopilot/context/glossary.md", "kind": "doc", "title": "Industry Glossary (External / Buyer-Facing)", "summary": "| Term | Definition |\n|------|-----------|\n| EPC | Engineering, Procurement, and Construction firm. Designs and builds capital projects. |\n| AEC | Architecture, Engineering, Construction. |", "path": "ops/autopilot/context/glossary.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/glossary.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "industry glossary (external / buyer-facing) | term | definition |\n|------|-----------|\n| epc | engineering, procurement, and construction firm. designs and builds capital projects. |\n| aec | architecture, engineering, construction. | ops/autopilot/context/glossary.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/glossary.md autopilot autopilot active docs autopilot context" }, { "id": "doc:docs/inspirations.md", "kind": "doc", "title": "Inspirations and Sources", "summary": "Non-dependency influences on Torus architecture. These projects shaped design decisions but don't appear in package.json. Review periodically for new ideas.\n---\n**Last reviewed:** 2026-05-21 (commit f2e11d6, v0.37.9.0)\n**Previously reviewed:** 2026-04-29 (commit 08746b0)", "path": "docs/inspirations.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/inspirations.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "inspirations and sources non-dependency influences on torus architecture. these projects shaped design decisions but don't appear in package.json. review periodically for new ideas.\n---\n**last reviewed:** 2026-05-21 (commit f2e11d6, v0.37.9.0)\n**previously reviewed:** 2026-04-29 (commit 08746b0) docs/inspirations.md https://github.com/torus-intelligence/torus/blob/staging/docs/inspirations.md github runbook active docs runbook docs" }, { "id": "doc:docs/superpowers/plans/2026-05-21-inspirations-borrows.md", "kind": "doc", "title": "Inspirations Borrows Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Land three borrows from the 2026-05-21 inspirations refresh: (B5) god-nodes + surprise scoring wired into the flag pipeline, (B4) cross-modal text↔image search via the existing sidecar, (B3) full hindsight calibration pipeline.\n**Architecture:** Each phase is independent and produces a working commit. B5 wires the already-built `findSurprises()` into the rule engine as observations and adds a PageRank centrality compute on top. B4 adds a multimodal embedding column on `documents` and a per-region table populated by a new `/embed/multimodal` endpoint on the exi...", "path": "docs/superpowers/plans/2026-05-21-inspirations-borrows.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-inspirations-borrows.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "inspirations borrows implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** land three borrows from the 2026-05-21 inspirations refresh: (b5) god-nodes + surprise scoring wired into the flag pipeline, (b4) cross-modal text↔image search via the existing sidecar, (b3) full hindsight calibration pipeline.\n**architecture:** each phase is independent and produces a working commit. b5 wires the already-built `findsurprises()` into the rule engine as observations and adds a pagerank centrality compute on top. b4 adds a multimodal embedding column on `documents` and a per-region table populated by a new `/embed/multimodal` endpoint on the exi... docs/superpowers/plans/2026-05-21-inspirations-borrows.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-inspirations-borrows.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/superpowers/specs/2026-05-21-inspirations-refresh-design.md", "kind": "doc", "title": "Inspirations Refresh -- 2026-05-21", "summary": "Survey of upstream activity across all active-influence sources since the 2026-04-29 review pass. Captures what's worth borrowing, with priority, rough size, and owner area. Items tagged \"now\" are mechanical bumps applied in the same commit as this spec.\nThree sources moved substantially: **gbrain** (25 minor versions, v0.22.10 -> v0.37.9, 105 commits), **pi-mono** (v0.73 -> v0.75, 389 commits), and **graphify** (v0.6.5 -> v0.8.14, 288 commits). One new source enters the catalog (**Mike**, AGPL legal doc platform -- design influence only). Source URL for OpenUI corrected (nichochar/open-ui is 404; active repo is thesysdev/openui). prosemirror-suggestion-mode repo deleted. DocDokuPLM confirmed dormant since 2021.\nOne mechanical bump lands in this pass: `@openuidev/svelte-lang` 0.1.1 -> 0.1...", "path": "docs/superpowers/specs/2026-05-21-inspirations-refresh-design.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-21-inspirations-refresh-design.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "inspirations refresh -- 2026-05-21 survey of upstream activity across all active-influence sources since the 2026-04-29 review pass. captures what's worth borrowing, with priority, rough size, and owner area. items tagged \"now\" are mechanical bumps applied in the same commit as this spec.\nthree sources moved substantially: **gbrain** (25 minor versions, v0.22.10 -> v0.37.9, 105 commits), **pi-mono** (v0.73 -> v0.75, 389 commits), and **graphify** (v0.6.5 -> v0.8.14, 288 commits). one new source enters the catalog (**mike**, agpl legal doc platform -- design influence only). source url for openui corrected (nichochar/open-ui is 404; active repo is thesysdev/openui). prosemirror-suggestion-mode repo deleted. docdokuplm confirmed dormant since 2021.\none mechanical bump lands in this pass: `@openuidev/svelte-lang` 0.1.1 -> 0.1... docs/superpowers/specs/2026-05-21-inspirations-refresh-design.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-21-inspirations-refresh-design.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:ops/autopilot/docs/integrations.md", "kind": "doc", "title": "Integrations Reference", "summary": "Reference for all external integrations available to autopilot skills.\nAttach these to your trigger's `mcp_connections` array:\n| Connector | UUID | What it does |\n|-----------|------|-------------|", "path": "ops/autopilot/docs/integrations.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/integrations.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "integrations reference reference for all external integrations available to autopilot skills.\nattach these to your trigger's `mcp_connections` array:\n| connector | uuid | what it does |\n|-----------|------|-------------| ops/autopilot/docs/integrations.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/integrations.md autopilot autopilot active docs autopilot docs" }, { "id": "skill:investor-update", "kind": "autopilot-skill", "title": "Investor Update", "summary": "---\nname: investor-update\ndescription: |\nEnd-of-month assembly. Aggregates the month's eow-review digests and current metrics into the investor update.", "path": "ops/autopilot/skills/investor-update/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/investor-update/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "investor-update" ], "searchText": "investor update ---\nname: investor-update\ndescription: |\nend-of-month assembly. aggregates the month's eow-review digests and current metrics into the investor update. ops/autopilot/skills/investor-update/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/investor-update/skill.md autopilot active autopilot skill skills investor-update" }, { "id": "doc:ops/autopilot/inventory/knowledge/README.md", "kind": "doc", "title": "Knowledge Base", "summary": "Accumulated learnings and signals from Torus agents and team members.\nEvery knowledge entry uses this frontmatter:\nDescribe the finding. Include:\n- **Context:** what was happening when this was discovered", "path": "ops/autopilot/inventory/knowledge/README.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/README.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "inventory", "knowledge" ], "searchText": "knowledge base accumulated learnings and signals from torus agents and team members.\nevery knowledge entry uses this frontmatter:\ndescribe the finding. include:\n- **context:** what was happening when this was discovered ops/autopilot/inventory/knowledge/readme.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/readme.md autopilot autopilot active docs autopilot inventory knowledge" }, { "id": "doc:docs/superpowers/specs/2026-05-20-lifecycle-audit-pod.md", "kind": "doc", "title": "Lifecycle audit pod — brief", "summary": "**Run date:** 2026-05-20 (this audit cycle scheduled the brief; the actual\naudit runs in a follow-up cycle).\n**Owner:** Tooling pod on next available audit week.\n`backend/lifecycle/*.ts`:", "path": "docs/superpowers/specs/2026-05-20-lifecycle-audit-pod.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-20-lifecycle-audit-pod.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "lifecycle audit pod — brief **run date:** 2026-05-20 (this audit cycle scheduled the brief; the actual\naudit runs in a follow-up cycle).\n**owner:** tooling pod on next available audit week.\n`backend/lifecycle/*.ts`: docs/superpowers/specs/2026-05-20-lifecycle-audit-pod.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-20-lifecycle-audit-pod.md github runbook active docs runbook docs superpowers specs" }, { "id": "autopilot-output:ops/autopilot/inventory/agent-state/linear-first-pass.json", "kind": "autopilot-output", "title": "Linear First Pass", "summary": "{\n \"last_run\": \"2026-04-03T18:00:00Z\",\n \"last_status\": \"NOTHING_TO_DO\",\n \"last_blocker\": null,\n \"last_blocker_posted\": null,\n \"last_blocker_first_seen\": null\n}", "path": "ops/autopilot/inventory/agent-state/linear-first-pass.json", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/linear-first-pass.json", "source": "AUTOPILOT", "category": "RUN_STATE", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "agent-state" ], "nextAction": "Review run state and resolve blocker if present.", "searchText": "linear first pass {\n \"last_run\": \"2026-04-03t18:00:00z\",\n \"last_status\": \"nothing_to_do\",\n \"last_blocker\": null,\n \"last_blocker_posted\": null,\n \"last_blocker_first_seen\": null\n} ops/autopilot/inventory/agent-state/linear-first-pass.json https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/linear-first-pass.json autopilot run_state needs-review review run state and resolve blocker if present. autopilot review inventory agent-state" }, { "id": "runner:linear-first-pass", "kind": "autopilot-runner", "title": "Linear First Pass", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/linear-first-pass/run.ts.", "path": "ops/autopilot/runners/linear-first-pass/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/linear-first-pass/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "linear-first-pass" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "linear first pass scheduled or operator-run autopilot workflow at ops/autopilot/runners/linear-first-pass/run.ts. ops/autopilot/runners/linear-first-pass/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/linear-first-pass/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner linear-first-pass" }, { "id": "skill:linear-first-pass", "kind": "autopilot-skill", "title": "Linear First-Pass", "summary": "---\nname: linear-first-pass\ndescription: |\nUse when a new Linear issue lands without the autopilot-reviewed label.", "path": "ops/autopilot/skills/linear-first-pass/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/linear-first-pass/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "linear-first-pass" ], "searchText": "linear first-pass ---\nname: linear-first-pass\ndescription: |\nuse when a new linear issue lands without the autopilot-reviewed label. ops/autopilot/skills/linear-first-pass/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/linear-first-pass/skill.md autopilot active autopilot skill skills linear-first-pass" }, { "id": "doc:ops/autopilot/context/linear.md", "kind": "doc", "title": "Linear Workspace", "summary": "- **Workspace:** torus\n- **URL:** https://linear.app/torus/\n| Team | ID | Subteams | Scope |\n|------|-----|----------|-------|", "path": "ops/autopilot/context/linear.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/linear.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "linear workspace - **workspace:** torus\n- **url:** https://linear.app/torus/\n| team | id | subteams | scope |\n|------|-----|----------|-------| ops/autopilot/context/linear.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/linear.md autopilot autopilot active docs autopilot context" }, { "id": "skill:linkedin-autoposter", "kind": "autopilot-skill", "title": "LinkedIn Autoposter", "summary": "---\nname: linkedin-autoposter\ndescription: |\nUse when managing the LinkedIn job posting automation service.", "path": "ops/autopilot/skills/linkedin-autoposter/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/linkedin-autoposter/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "linkedin-autoposter" ], "searchText": "linkedin autoposter ---\nname: linkedin-autoposter\ndescription: |\nuse when managing the linkedin job posting automation service. ops/autopilot/skills/linkedin-autoposter/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/linkedin-autoposter/skill.md autopilot active autopilot skill skills linkedin-autoposter" }, { "id": "skill:linkedin-performance", "kind": "autopilot-skill", "title": "LinkedIn Performance Analyzer", "summary": "---\nname: linkedin-performance\ndescription: |\nUse when analyzing LinkedIn post performance from screenshots or manually provided metrics.", "path": "ops/autopilot/skills/linkedin-performance/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/linkedin-performance/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "linkedin-performance" ], "searchText": "linkedin performance analyzer ---\nname: linkedin-performance\ndescription: |\nuse when analyzing linkedin post performance from screenshots or manually provided metrics. ops/autopilot/skills/linkedin-performance/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/linkedin-performance/skill.md autopilot active autopilot skill skills linkedin-performance" }, { "id": "doc:ops/autopilot/context/models.md", "kind": "doc", "title": "LLM / ML Model Catalog", "summary": "| Tier | Provider | Model | Use |\n|------|----------|-------|-----|\n| Opus | Anthropic | claude-opus-4-7 | Agent planning, complex reasoning |\n| Sonnet | Anthropic | claude-sonnet-4-6 | Default agent work |", "path": "ops/autopilot/context/models.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/models.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "llm / ml model catalog | tier | provider | model | use |\n|------|----------|-------|-----|\n| opus | anthropic | claude-opus-4-7 | agent planning, complex reasoning |\n| sonnet | anthropic | claude-sonnet-4-6 | default agent work | ops/autopilot/context/models.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/models.md autopilot autopilot active docs autopilot context" }, { "id": "doc:docs/superpowers/specs/2026-05-20-magic-link-audit.md", "kind": "doc", "title": "Magic-link audit — brief", "summary": "**Run date:** 2026-05-20 (this audit cycle scheduled the brief; the actual\naudit runs in a follow-up cycle).\n**Owner:** Auth pod on next available audit week.\n- `frontend/src/routes/login/verify/+page.server.ts` — the verification", "path": "docs/superpowers/specs/2026-05-20-magic-link-audit.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-20-magic-link-audit.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "magic-link audit — brief **run date:** 2026-05-20 (this audit cycle scheduled the brief; the actual\naudit runs in a follow-up cycle).\n**owner:** auth pod on next available audit week.\n- `frontend/src/routes/login/verify/+page.server.ts` — the verification docs/superpowers/specs/2026-05-20-magic-link-audit.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-20-magic-link-audit.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:docs/superpowers/plans/2026-05-21-markdown-engine-coexistence.md", "kind": "doc", "title": "Markdown engine coexistence note (W2-23 follow-up)", "summary": "**Status:** Documentation-only. No code changes implied.\nTwo markdown engines live in `frontend/`:\n| Engine | Where | Why |\n|--------|-------|-----|", "path": "docs/superpowers/plans/2026-05-21-markdown-engine-coexistence.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-markdown-engine-coexistence.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "markdown engine coexistence note (w2-23 follow-up) **status:** documentation-only. no code changes implied.\ntwo markdown engines live in `frontend/`:\n| engine | where | why |\n|--------|-------|-----| docs/superpowers/plans/2026-05-21-markdown-engine-coexistence.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-markdown-engine-coexistence.md github runbook active docs runbook docs superpowers plans" }, { "id": "runner:marketing-deep-dive", "kind": "autopilot-runner", "title": "Marketing Deep Dive", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/marketing-deep-dive/run.ts.", "path": "ops/autopilot/runners/marketing-deep-dive/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/marketing-deep-dive/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "marketing-deep-dive" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "marketing deep dive scheduled or operator-run autopilot workflow at ops/autopilot/runners/marketing-deep-dive/run.ts. ops/autopilot/runners/marketing-deep-dive/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/marketing-deep-dive/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner marketing-deep-dive" }, { "id": "skill:marketing-deep-dive", "kind": "autopilot-skill", "title": "Marketing Deep Dive", "summary": "---\nname: marketing-deep-dive\ndescription: Saturday weekly marketing/content review. 8 sub-agents. Continue-on-failure with per-sub-agent time budget.\n---", "path": "ops/autopilot/skills/marketing-deep-dive/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/marketing-deep-dive/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "marketing-deep-dive" ], "searchText": "marketing deep dive ---\nname: marketing-deep-dive\ndescription: saturday weekly marketing/content review. 8 sub-agents. continue-on-failure with per-sub-agent time budget.\n--- ops/autopilot/skills/marketing-deep-dive/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/marketing-deep-dive/skill.md autopilot active autopilot skill skills marketing-deep-dive" }, { "id": "skill:meeting-notes-to-linear", "kind": "autopilot-skill", "title": "Meeting Notes to Linear", "summary": "---\nname: meeting-notes-to-linear\ndescription: |\nUse when processing meeting transcripts from Granola into actionable Linear tickets.", "path": "ops/autopilot/skills/meeting-notes-to-linear/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/meeting-notes-to-linear/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "meeting-notes-to-linear" ], "searchText": "meeting notes to linear ---\nname: meeting-notes-to-linear\ndescription: |\nuse when processing meeting transcripts from granola into actionable linear tickets. ops/autopilot/skills/meeting-notes-to-linear/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/meeting-notes-to-linear/skill.md autopilot active autopilot skill skills meeting-notes-to-linear" }, { "id": "autopilot-output:ops/autopilot/inventory/agent-state/meeting-notes-to-linear.json", "kind": "autopilot-output", "title": "Meeting Notes To Linear", "summary": "{\n \"last_run\": \"2026-04-03T12:00:00Z\",\n \"last_status\": \"DONE_EMPTY_NO_MEETINGS\",\n \"last_blocker\": null,\n \"last_blocker_posted\": null,\n \"last_blocker_first_seen\": null\n}", "path": "ops/autopilot/inventory/agent-state/meeting-notes-to-linear.json", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/meeting-notes-to-linear.json", "source": "AUTOPILOT", "category": "RUN_STATE", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "agent-state" ], "nextAction": "Review run state and resolve blocker if present.", "searchText": "meeting notes to linear {\n \"last_run\": \"2026-04-03t12:00:00z\",\n \"last_status\": \"done_empty_no_meetings\",\n \"last_blocker\": null,\n \"last_blocker_posted\": null,\n \"last_blocker_first_seen\": null\n} ops/autopilot/inventory/agent-state/meeting-notes-to-linear.json https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/agent-state/meeting-notes-to-linear.json autopilot run_state needs-review review run state and resolve blocker if present. autopilot review inventory agent-state" }, { "id": "doc:docs/superpowers/specs/2026-04-21-mega-audit-design.md", "kind": "doc", "title": "Mega-Audit of Legend — Design Spec", "summary": "**Date:** 2026-04-21\n**Branch:** staging (full repo, all surfaces)\n**Output:** One mega-report + master overhaul plan + Tier-1 commit branch\n**Target runtime:** ~3–4 hours overnight", "path": "docs/superpowers/specs/2026-04-21-mega-audit-design.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-04-21-mega-audit-design.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "mega-audit of legend — design spec **date:** 2026-04-21\n**branch:** staging (full repo, all surfaces)\n**output:** one mega-report + master overhaul plan + tier-1 commit branch\n**target runtime:** ~3–4 hours overnight docs/superpowers/specs/2026-04-21-mega-audit-design.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-04-21-mega-audit-design.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:docs/superpowers/plans/2026-04-21-mega-audit-of-legend.md", "kind": "doc", "title": "Mega-Audit of Legend — Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Run a comprehensive overnight overhaul of the entire Torus repo across 34 lenses organized into 7 hybrid pods, producing a mega-report, a master overhaul plan, and a `mega-audit-fixes` branch with Tier-1 mechanical fixes committed.\n**Architecture:** Orchestrator (main session) dispatches pod agents in phases. Analysis pods 1–6 run in parallel (isolated worktrees, read-only). Pod 7 runs after, then merger, then master-plan, then Tier-1 executor on a fresh writable worktree. Each pod agent receives a full briefing with mandate, scope, output contract, tier rules...", "path": "docs/superpowers/plans/2026-04-21-mega-audit-of-legend.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-04-21-mega-audit-of-legend.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "mega-audit of legend — implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** run a comprehensive overnight overhaul of the entire torus repo across 34 lenses organized into 7 hybrid pods, producing a mega-report, a master overhaul plan, and a `mega-audit-fixes` branch with tier-1 mechanical fixes committed.\n**architecture:** orchestrator (main session) dispatches pod agents in phases. analysis pods 1–6 run in parallel (isolated worktrees, read-only). pod 7 runs after, then merger, then master-plan, then tier-1 executor on a fresh writable worktree. each pod agent receives a full briefing with mandate, scope, output contract, tier rules... docs/superpowers/plans/2026-04-21-mega-audit-of-legend.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-04-21-mega-audit-of-legend.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/inspirations/mike.md", "kind": "doc", "title": "Mike (willchen96/mike)", "summary": "**Source:** https://github.com/willchen96/mike -- https://mikeoss.com\n**License:** AGPL-3.0 (read for inspiration only; no code copy)\n**First documented:** 2026-05-21\n**Last reviewed:** 2026-05-21 (commit 4f33843)", "path": "docs/inspirations/mike.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/inspirations/mike.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "inspirations" ], "searchText": "mike (willchen96/mike) **source:** https://github.com/willchen96/mike -- https://mikeoss.com\n**license:** agpl-3.0 (read for inspiration only; no code copy)\n**first documented:** 2026-05-21\n**last reviewed:** 2026-05-21 (commit 4f33843) docs/inspirations/mike.md https://github.com/torus-intelligence/torus/blob/staging/docs/inspirations/mike.md github runbook active docs runbook docs inspirations" }, { "id": "doc:docs/superpowers/specs/2026-05-05-mike-superset-absorption-roadmap.md", "kind": "doc", "title": "Mike + Superset Absorption Roadmap", "summary": "**Date:** 2026-05-05\n**Status:** Draft (meta-roadmap, no implementation detail)\n**Scope:** Cross-cutting. Ranks and orders 9 work buckets extracted from willchen96/mike + superset-sh/superset by dependency. No calendar — buckets ship in dependency order, in parallel where possible.\n**Companion:** `docs/inspirations.md` will get Mike + Superset entries in the canonical \"What we took / What we skipped / Review for\" format as a separate, smaller pass.", "path": "docs/superpowers/specs/2026-05-05-mike-superset-absorption-roadmap.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-05-mike-superset-absorption-roadmap.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "mike + superset absorption roadmap **date:** 2026-05-05\n**status:** draft (meta-roadmap, no implementation detail)\n**scope:** cross-cutting. ranks and orders 9 work buckets extracted from willchen96/mike + superset-sh/superset by dependency. no calendar — buckets ship in dependency order, in parallel where possible.\n**companion:** `docs/inspirations.md` will get mike + superset entries in the canonical \"what we took / what we skipped / review for\" format as a separate, smaller pass. docs/superpowers/specs/2026-05-05-mike-superset-absorption-roadmap.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-05-05-mike-superset-absorption-roadmap.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:ops/autopilot/context/targets.md", "kind": "doc", "title": "Named Target Accounts", "summary": "`[PENDING: pull from current pursuits + ENR Top 400 + named verticals in CLAUDE.md]`\n| Firm | Tier | Vertical | Region | Status | Contact | Last touch |\n|------|------|----------|--------|--------|---------|------------|\n- LNG (export terminals, gas processing)", "path": "ops/autopilot/context/targets.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/targets.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "named target accounts `[pending: pull from current pursuits + enr top 400 + named verticals in claude.md]`\n| firm | tier | vertical | region | status | contact | last touch |\n|------|------|----------|--------|--------|---------|------------|\n- lng (export terminals, gas processing) ops/autopilot/context/targets.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/targets.md autopilot autopilot active docs autopilot context" }, { "id": "doc:docs/ops-console.md", "kind": "doc", "title": "Ops Console", "summary": "`ops.usetorus.com` is the internal Torus operations surface. It should not be a\nsingle embedded product or a dumping ground. It should be a small, controlled\nportal over operational systems that humans and agents can query through the\nsame permission model.", "path": "docs/ops-console.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/ops-console.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "ops console `ops.usetorus.com` is the internal torus operations surface. it should not be a\nsingle embedded product or a dumping ground. it should be a small, controlled\nportal over operational systems that humans and agents can query through the\nsame permission model. docs/ops-console.md https://github.com/torus-intelligence/torus/blob/staging/docs/ops-console.md github runbook active docs runbook docs" }, { "id": "doc:docs/ops-crm-workflows.md", "kind": "doc", "title": "Ops CRM Workflows", "summary": "Twenty is the CRM module inside Ops, not the whole Ops product. Keep source\ntruth in GitHub, Autopilot, Granola, Railway, and Torus services; sync compact,\nattributed records into Twenty so humans and agents can inspect and act.\n- `ops/crm/app`: private Twenty app metadata for Torus-specific objects,", "path": "docs/ops-crm-workflows.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/ops-crm-workflows.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "ops crm workflows twenty is the crm module inside ops, not the whole ops product. keep source\ntruth in github, autopilot, granola, railway, and torus services; sync compact,\nattributed records into twenty so humans and agents can inspect and act.\n- `ops/crm/app`: private twenty app metadata for torus-specific objects, docs/ops-crm-workflows.md https://github.com/torus-intelligence/torus/blob/staging/docs/ops-crm-workflows.md github runbook active docs runbook docs" }, { "id": "doc:ops/autopilot/context/outbound.md", "kind": "doc", "title": "Outbound Sequences -- EPC Personas", "summary": "5-step touch cadence per persona. Templates live at `ops/autopilot/skills/cold-outreach/templates/`. Day numbers are days-since-first-touch.\n- D0 -- intro + specific cross-discipline catch (cold-1)\n- D3 -- thread reply with second angle (cold-2)\n- D7 -- LinkedIn manual send (cold-3)", "path": "ops/autopilot/context/outbound.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/outbound.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "outbound sequences -- epc personas 5-step touch cadence per persona. templates live at `ops/autopilot/skills/cold-outreach/templates/`. day numbers are days-since-first-touch.\n- d0 -- intro + specific cross-discipline catch (cold-1)\n- d3 -- thread reply with second angle (cold-2)\n- d7 -- linkedin manual send (cold-3) ops/autopilot/context/outbound.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/outbound.md autopilot autopilot active docs autopilot context" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/attribution-swaps.md", "kind": "doc", "title": "Parallel-Subagent Attribution Swaps — 2026-05-21 Weekly Audit", "summary": "During the Wave 4 security cluster dispatch (batch 10, 14 parallel subagents\ntouching disjoint files), `git add` / `git commit` calls across subagents\ninterleaved in the shared worktree. Pathspec discipline prevented file\ncorruption — every subagent's CODE landed correctly — but four commits ended", "path": "docs/superpowers/audits/2026-05-21-weekly/attribution-swaps.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/attribution-swaps.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "parallel-subagent attribution swaps — 2026-05-21 weekly audit during the wave 4 security cluster dispatch (batch 10, 14 parallel subagents\ntouching disjoint files), `git add` / `git commit` calls across subagents\ninterleaved in the shared worktree. pathspec discipline prevented file\ncorruption — every subagent's code landed correctly — but four commits ended docs/superpowers/audits/2026-05-21-weekly/attribution-swaps.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/attribution-swaps.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/people-and-organizations.md", "kind": "doc", "title": "People and Organizations", "summary": "Plugin: `backend/plugins/people/`. Agent access via the consolidated `search` tool (`backend/plugins/agent/tools/search.ts`, scope=\"people\").\nExtracts, reconciles, and maintains a graph of people and organizations from emails, documents, and manual imports.\n- canonical_name, emails[]\n- sources: email | document | roster | manual", "path": "docs/people-and-organizations.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/people-and-organizations.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "people and organizations plugin: `backend/plugins/people/`. agent access via the consolidated `search` tool (`backend/plugins/agent/tools/search.ts`, scope=\"people\").\nextracts, reconciles, and maintains a graph of people and organizations from emails, documents, and manual imports.\n- canonical_name, emails[]\n- sources: email | document | roster | manual docs/people-and-organizations.md https://github.com/torus-intelligence/torus/blob/staging/docs/people-and-organizations.md github runbook active docs runbook docs" }, { "id": "doc:docs/superpowers/plans/INDEX.md", "kind": "doc", "title": "Plans Index", "summary": "One-line summary of every plan in `docs/superpowers/plans/`.\n- [2026-04-21-mega-audit-of-legend.md](./2026-04-21-mega-audit-of-legend.md) — Overnight overhaul plan: 34-lens audit across 7 pods, producing mega-report, master overhaul plan, and a `mega-audit-fixes` branch with Tier-1 mechanical fixes.\n- [2026-04-22-recurring-audit.md](./2026-04-22-recurring-audit.md) — Recurring audit plan running the mega-audit in weekly (full sweep) or nightly (diff-scoped) mode with date-stamped outputs.\n- [2026-05-20-weekly-overhaul.md](./2026-05-20-weekly-overhaul.md) — Weekly overhaul plan derived from the 2026-05-20 v2 deep audit (246 findings post-dedup).", "path": "docs/superpowers/plans/INDEX.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/INDEX.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "plans index one-line summary of every plan in `docs/superpowers/plans/`.\n- [2026-04-21-mega-audit-of-legend.md](./2026-04-21-mega-audit-of-legend.md) — overnight overhaul plan: 34-lens audit across 7 pods, producing mega-report, master overhaul plan, and a `mega-audit-fixes` branch with tier-1 mechanical fixes.\n- [2026-04-22-recurring-audit.md](./2026-04-22-recurring-audit.md) — recurring audit plan running the mega-audit in weekly (full sweep) or nightly (diff-scoped) mode with date-stamped outputs.\n- [2026-05-20-weekly-overhaul.md](./2026-05-20-weekly-overhaul.md) — weekly overhaul plan derived from the 2026-05-20 v2 deep audit (246 findings post-dedup). docs/superpowers/plans/index.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/index.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:ops/autopilot/docs/platform.md", "kind": "doc", "title": "Platform -- What You Can Build", "summary": "The platform is Torus's data backbone. It stores structured data, receives webhooks, and serves queries. You don't need to know how it works -- just tell Claude what you need.\nExamples:\n- \"I want to collect feedback from our event booths\"\n- \"I want to track which prospects we've contacted\"", "path": "ops/autopilot/docs/platform.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/docs/platform.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "docs" ], "searchText": "platform -- what you can build the platform is torus's data backbone. it stores structured data, receives webhooks, and serves queries. you don't need to know how it works -- just tell claude what you need.\nexamples:\n- \"i want to collect feedback from our event booths\"\n- \"i want to track which prospects we've contacted\" ops/autopilot/docs/platform.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/docs/platform.md autopilot autopilot active docs autopilot docs" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-1-security.md", "kind": "doc", "title": "Pod 1 (Security) — Weekly save-point audit, 2026-05-21", "summary": "**Pod:** Pod 1 (Security)\n**Lenses covered:** 4 (CSO comprehensive), 31 (agent red-team), 24 (secrets & env config), 34 (license & attribution)\n**Mode:** weekly save-point — read-only static analysis at staging tip `8125d2f1`\n**Context:** PR #44 (merged 2026-05-20) shipped Wave 0-4 remediations of last week's findings. This audit verifies staging is clean and re-flags anything not actually closed.", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-1-security.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-1-security.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 1 (security) — weekly save-point audit, 2026-05-21 **pod:** pod 1 (security)\n**lenses covered:** 4 (cso comprehensive), 31 (agent red-team), 24 (secrets & env config), 34 (license & attribution)\n**mode:** weekly save-point — read-only static analysis at staging tip `8125d2f1`\n**context:** pr #44 (merged 2026-05-20) shipped wave 0-4 remediations of last week's findings. this audit verifies staging is clean and re-flags anything not actually closed. docs/superpowers/audits/2026-05-21-weekly/pod-1-security.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-1-security.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-1-security.md", "kind": "doc", "title": "Pod 1 v2 (deep) — Security", "summary": "**Lenses covered:** 4 (CSO comprehensive), 31 (agent red-team), 24 (secrets & env config), 34 (license & attribution)\n**Run:** 2026-05-20 weekly mega-audit, v2 deep pass\n**Mode:** read-only static analysis, exhaustive route-by-route + boundary-by-boundary\n**42 findings total: 7 critical, 13 high, 14 medium, 6 low, 2 nit**", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-1-security.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-1-security.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 1 v2 (deep) — security **lenses covered:** 4 (cso comprehensive), 31 (agent red-team), 24 (secrets & env config), 34 (license & attribution)\n**run:** 2026-05-20 weekly mega-audit, v2 deep pass\n**mode:** read-only static analysis, exhaustive route-by-route + boundary-by-boundary\n**42 findings total: 7 critical, 13 high, 14 medium, 6 low, 2 nit** docs/superpowers/audits/2026-05-20-weekly/pod-1-security.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-1-security.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-2-correctness.md", "kind": "doc", "title": "Pod 2 — Correctness (2026-05-21 weekly)", "summary": "**Lenses covered:** 6 (type safety), 17 (error handling & resilience), 20 (concurrency & races), 18 (data integrity), 28 (plan runner durability)\n**Mode:** weekly full sweep, read-only static analysis. Staging tip 8125d2f1.\n**Worktree:** `/Users/rahulthayil/.superset/worktrees/torus/spangle-flame`\n**11 findings: 0 critical, 6 high, 3 medium, 2 low, 0 nit**", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-2-correctness.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-2-correctness.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 2 — correctness (2026-05-21 weekly) **lenses covered:** 6 (type safety), 17 (error handling & resilience), 20 (concurrency & races), 18 (data integrity), 28 (plan runner durability)\n**mode:** weekly full sweep, read-only static analysis. staging tip 8125d2f1.\n**worktree:** `/users/rahulthayil/.superset/worktrees/torus/spangle-flame`\n**11 findings: 0 critical, 6 high, 3 medium, 2 low, 0 nit** docs/superpowers/audits/2026-05-21-weekly/pod-2-correctness.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-2-correctness.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-2-correctness.md", "kind": "doc", "title": "Pod 2 v2 (deep) — Correctness", "summary": "**Lenses covered:** 6 (type safety), 17 (error handling / resilience), 20 (concurrency & races), 18 (data integrity), 28 (plan runner durability)\n**Run:** 2026-05-20 weekly mega-audit, v2 (deep) pass.\n**Mode:** read-only static analysis on the orchestrator worktree\n(`/Users/rahulthayil/.superset/worktrees/torus/pleasant-list`) plus targeted", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-2-correctness.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-2-correctness.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 2 v2 (deep) — correctness **lenses covered:** 6 (type safety), 17 (error handling / resilience), 20 (concurrency & races), 18 (data integrity), 28 (plan runner durability)\n**run:** 2026-05-20 weekly mega-audit, v2 (deep) pass.\n**mode:** read-only static analysis on the orchestrator worktree\n(`/users/rahulthayil/.superset/worktrees/torus/pleasant-list`) plus targeted docs/superpowers/audits/2026-05-20-weekly/pod-2-correctness.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-2-correctness.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md", "kind": "doc", "title": "Pod 3 — Quality (v2 deep)", "summary": "**Lenses:** 2 (dead code & duplication), 3 (centralization), 14 (simplification), 32 (Chekhov's gun), 33 (naming & domain vocabulary)\n**Run:** 2026-05-20 weekly mega-audit, DEEP pass\n**Mode:** read-only, parallel sub-fanout (6 sub-agents), main worktree\n**35 findings: 0 critical, 7 high, 19 medium, 7 low, 2 nit**", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 3 — quality (v2 deep) **lenses:** 2 (dead code & duplication), 3 (centralization), 14 (simplification), 32 (chekhov's gun), 33 (naming & domain vocabulary)\n**run:** 2026-05-20 weekly mega-audit, deep pass\n**mode:** read-only, parallel sub-fanout (6 sub-agents), main worktree\n**35 findings: 0 critical, 7 high, 19 medium, 7 low, 2 nit** docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-3-quality.md", "kind": "doc", "title": "Pod 3 — Quality (weekly 2026-05-21)", "summary": "**Lenses:** 2 (dead code & duplication), 3 (centralization), 14 (simplification), 32 (Chekhov's gun), 33 (naming & domain vocabulary)\n**Mode:** weekly (full sweep), read-only, main worktree, staging tip `8125d2f1`\n**Scope:** engineering surfaces (backend/, frontend/, website/, sidecar/, deploy/). Autopilot is out of scope per CLAUDE.md.\n**Prior pod:** `docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md` (\"v2 deep\", 35 findings, FINDING-pod-3-v2-{1..35})", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-3-quality.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-3-quality.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 3 — quality (weekly 2026-05-21) **lenses:** 2 (dead code & duplication), 3 (centralization), 14 (simplification), 32 (chekhov's gun), 33 (naming & domain vocabulary)\n**mode:** weekly (full sweep), read-only, main worktree, staging tip `8125d2f1`\n**scope:** engineering surfaces (backend/, frontend/, website/, sidecar/, deploy/). autopilot is out of scope per claude.md.\n**prior pod:** `docs/superpowers/audits/2026-05-20-weekly/pod-3-quality.md` (\"v2 deep\", 35 findings, finding-pod-3-v2-{1..35}) docs/superpowers/audits/2026-05-21-weekly/pod-3-quality.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-3-quality.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-4-architecture.md", "kind": "doc", "title": "Pod 4 — Architecture (weekly mega-audit 2026-05-21)", "summary": "**Pod:** pod-4 Architecture\n**Lenses covered:** 1 (Architecture & boundaries), 10 (DB schema health), 19 (API contract hygiene), 27 (KG integrity), 26 (Email/DNS pipeline)\n**Mode:** weekly full sweep, read-only\n**Staging tip:** 8125d2f1 (post-PR #44, post wave-0..4 merge, post deploy fixups)", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-4-architecture.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-4-architecture.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 4 — architecture (weekly mega-audit 2026-05-21) **pod:** pod-4 architecture\n**lenses covered:** 1 (architecture & boundaries), 10 (db schema health), 19 (api contract hygiene), 27 (kg integrity), 26 (email/dns pipeline)\n**mode:** weekly full sweep, read-only\n**staging tip:** 8125d2f1 (post-pr #44, post wave-0..4 merge, post deploy fixups) docs/superpowers/audits/2026-05-21-weekly/pod-4-architecture.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-4-architecture.md github architecture active docs architecture docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-4-architecture.md", "kind": "doc", "title": "Pod 4 — Architecture v2 (DEEP) — weekly mega-audit 2026-05-20", "summary": "**Pod:** pod-4 Architecture (DEEP pass)\n**Lenses covered:** 1 (Architecture & boundaries), 10 (DB schema health), 19 (API contract hygiene), 27 (KG integrity), 26 (Email/DNS pipeline)\n**Run mode:** weekly deep sweep, read-only, no token/time cap\n**Inputs:** v1 (16 findings), pod-7 adversarial (28 findings + contradictions), CLAUDE.md, db/schema.md, db/schema.condensed.md, all 89 migrations, all backend plugins", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-4-architecture.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-4-architecture.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 4 — architecture v2 (deep) — weekly mega-audit 2026-05-20 **pod:** pod-4 architecture (deep pass)\n**lenses covered:** 1 (architecture & boundaries), 10 (db schema health), 19 (api contract hygiene), 27 (kg integrity), 26 (email/dns pipeline)\n**run mode:** weekly deep sweep, read-only, no token/time cap\n**inputs:** v1 (16 findings), pod-7 adversarial (28 findings + contradictions), claude.md, db/schema.md, db/schema.condensed.md, all 89 migrations, all backend plugins docs/superpowers/audits/2026-05-20-weekly/pod-4-architecture.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-4-architecture.md github architecture active docs architecture docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-5-performance.md", "kind": "doc", "title": "Pod 5 — Performance", "summary": "**Audit:** weekly (full sweep) — 2026-05-21\n**Branch tip:** `8125d2f1` (spangle-flame, mirrors staging)\n**Scope:** backend/plugins/* (DB-touching), frontend/src/*, backend/plugins/llm/*, backend/storage.ts, .github/workflows/*, turbo.json\n**Lenses owned:** 5 (perf hotspots) → 21 (LLM cost & token waste) → 29 (storage hygiene) → 25 (build & CI) → 30 (frontend state & SSR)", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-5-performance.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-5-performance.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 5 — performance **audit:** weekly (full sweep) — 2026-05-21\n**branch tip:** `8125d2f1` (spangle-flame, mirrors staging)\n**scope:** backend/plugins/* (db-touching), frontend/src/*, backend/plugins/llm/*, backend/storage.ts, .github/workflows/*, turbo.json\n**lenses owned:** 5 (perf hotspots) → 21 (llm cost & token waste) → 29 (storage hygiene) → 25 (build & ci) → 30 (frontend state & ssr) docs/superpowers/audits/2026-05-21-weekly/pod-5-performance.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-5-performance.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-5-performance.md", "kind": "doc", "title": "Pod 5 v2 (deep) — Performance", "summary": "**Lenses covered:** 5 (Performance hotspots), 21 (LLM cost & token waste), 29 (Storage hygiene), 25 (Build & CI), 30 (Frontend state & SSR)\n**Run date:** 2026-05-20 (weekly DEEP v2 pass; no token/time cap)\n**Mode:** read-only, parallel sub-fanout (sub-5a DB, sub-5b LLM cost, sub-5c storage, sub-5d build/CI, sub-5e frontend)\n---", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-5-performance.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-5-performance.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 5 v2 (deep) — performance **lenses covered:** 5 (performance hotspots), 21 (llm cost & token waste), 29 (storage hygiene), 25 (build & ci), 30 (frontend state & ssr)\n**run date:** 2026-05-20 (weekly deep v2 pass; no token/time cap)\n**mode:** read-only, parallel sub-fanout (sub-5a db, sub-5b llm cost, sub-5c storage, sub-5d build/ci, sub-5e frontend)\n--- docs/superpowers/audits/2026-05-20-weekly/pod-5-performance.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-5-performance.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-6-ux-docs.md", "kind": "doc", "title": "Pod 6 — UX & Docs (weekly, 2026-05-21)", "summary": "**Mode:** weekly (full sweep)\n**Worktree tip:** 8125d2f1\n**Scope:** Lenses 8 (frontend design/UX), 9 (DevEx), 22 (accessibility), 23 (content/tone), 12 (docs staleness).\n**Method:** Read-only Grep/Read sweep across `frontend/src`, `website/src`, `ops/autopilot/`, `docs/`, `CLAUDE.md`, `DESIGN.md`, `README.md`, `CONTRIBUTING.md`, `db/schema.md`, `backend/plugins/agent/system-prompt.ts`, plan dir, plus targeted verification of every prior pod-6/pod-7 v2 finding called out in `2026-05-20-weekly.md`.", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-6-ux-docs.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-6-ux-docs.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 6 — ux & docs (weekly, 2026-05-21) **mode:** weekly (full sweep)\n**worktree tip:** 8125d2f1\n**scope:** lenses 8 (frontend design/ux), 9 (devex), 22 (accessibility), 23 (content/tone), 12 (docs staleness).\n**method:** read-only grep/read sweep across `frontend/src`, `website/src`, `ops/autopilot/`, `docs/`, `claude.md`, `design.md`, `readme.md`, `contributing.md`, `db/schema.md`, `backend/plugins/agent/system-prompt.ts`, plan dir, plus targeted verification of every prior pod-6/pod-7 v2 finding called out in `2026-05-20-weekly.md`. docs/superpowers/audits/2026-05-21-weekly/pod-6-ux-docs.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-6-ux-docs.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-6-ux-docs.md", "kind": "doc", "title": "Pod 6 v2 (deep) — UX & Docs", "summary": "**Lenses covered:** 8 (Frontend design & UX), 9 (DevEx), 22 (Accessibility deep), 23 (Content & tone), 12 (Docs staleness)\n**Run:** 2026-05-20 weekly mega-audit, deep pass\n**Mode:** read-only, no commits\n**54 findings: 1 critical, 14 high, 21 medium, 13 low, 5 nit**", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-6-ux-docs.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-6-ux-docs.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 6 v2 (deep) — ux & docs **lenses covered:** 8 (frontend design & ux), 9 (devex), 22 (accessibility deep), 23 (content & tone), 12 (docs staleness)\n**run:** 2026-05-20 weekly mega-audit, deep pass\n**mode:** read-only, no commits\n**54 findings: 1 critical, 14 high, 21 medium, 13 low, 5 nit** docs/superpowers/audits/2026-05-20-weekly/pod-6-ux-docs.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-6-ux-docs.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/pod-7-adversarial.md", "kind": "doc", "title": "Pod 7 — Adversarial (weekly 2026-05-21)", "summary": "**Pod:** Pod 7 (Adversarial)\n**Lenses covered:** 13 (Codex challenge on critical modules), 11 (Agent/LLM quality), 7 (Test coverage & quality), 15 (Dependency hygiene)\n**Mode:** weekly full sweep, read-only static analysis at staging tip `8125d2f1`\n**Method:** Independent verification of pods 1-6 plus adversarial reads of the 10 highest-trust modules (trust filter, approval state, PlanRunner, withFirmContext, email router, workspace generation, agent tool registry, consistency rules, KG union views, lifecycle/graduation).", "path": "docs/superpowers/audits/2026-05-21-weekly/pod-7-adversarial.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-7-adversarial.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "pod 7 — adversarial (weekly 2026-05-21) **pod:** pod 7 (adversarial)\n**lenses covered:** 13 (codex challenge on critical modules), 11 (agent/llm quality), 7 (test coverage & quality), 15 (dependency hygiene)\n**mode:** weekly full sweep, read-only static analysis at staging tip `8125d2f1`\n**method:** independent verification of pods 1-6 plus adversarial reads of the 10 highest-trust modules (trust filter, approval state, planrunner, withfirmcontext, email router, workspace generation, agent tool registry, consistency rules, kg union views, lifecycle/graduation). docs/superpowers/audits/2026-05-21-weekly/pod-7-adversarial.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/pod-7-adversarial.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/pod-7-adversarial.md", "kind": "doc", "title": "Pod 7 v2 (deep adversarial) — 2 rounds + gaps", "summary": "**Lenses covered:** 13 (Codex challenge on top-10 critical modules), 11 (Agent/LLM quality), 7 (Test coverage & quality), 15 (Dependency hygiene); plus a cross-pod stress-test of v2 pods 1-6 and a gap sweep against the spec's 34 lenses.\n**Run:** 2026-05-20 weekly mega-audit, v2 (deep) adversarial pass\n**Mode:** read-only static analysis on the orchestrator worktree at `/Users/rahulthayil/.superset/worktrees/torus/pleasant-list`. Verified critical claims by re-reading file:lines, running `bunx tsc --noEmit` (backend, EXIT=0), `bunx svelte-check --workspace frontend` (1292 files, 0 errors, 1 warning), `bunx biome check . --reporter=summary` (2 warnings + 1 info), and querying `registry.npmjs.org` + `pypi.org` directly for phantom-version claims.\n**52 findings: 5 critical, 19 high, 19 medium...", "path": "docs/superpowers/audits/2026-05-20-weekly/pod-7-adversarial.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-7-adversarial.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "pod 7 v2 (deep adversarial) — 2 rounds + gaps **lenses covered:** 13 (codex challenge on top-10 critical modules), 11 (agent/llm quality), 7 (test coverage & quality), 15 (dependency hygiene); plus a cross-pod stress-test of v2 pods 1-6 and a gap sweep against the spec's 34 lenses.\n**run:** 2026-05-20 weekly mega-audit, v2 (deep) adversarial pass\n**mode:** read-only static analysis on the orchestrator worktree at `/users/rahulthayil/.superset/worktrees/torus/pleasant-list`. verified critical claims by re-reading file:lines, running `bunx tsc --noemit` (backend, exit=0), `bunx svelte-check --workspace frontend` (1292 files, 0 errors, 1 warning), `bunx biome check . --reporter=summary` (2 warnings + 1 info), and querying `registry.npmjs.org` + `pypi.org` directly for phantom-version claims.\n**52 findings: 5 critical, 19 high, 19 medium... docs/superpowers/audits/2026-05-20-weekly/pod-7-adversarial.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/pod-7-adversarial.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/posthog-setup.md", "kind": "doc", "title": "PostHog setup", "summary": "Everything is wired in code across the three surfaces (marketing Astro, SvelteKit app, Bun backend). Every helper no-ops when the keys are unset, so sovereign installs leak nothing. This doc is the one-time setup to light it up.\n1. Sign up at [posthog.com](https://posthog.com), pick **Cloud-US**\n2. Create one project: `Torus`\n3. Copy the **Project API Key** (`phc_xxxxxxxxxxxxx`) from Project Settings", "path": "docs/posthog-setup.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/posthog-setup.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "posthog setup everything is wired in code across the three surfaces (marketing astro, sveltekit app, bun backend). every helper no-ops when the keys are unset, so sovereign installs leak nothing. this doc is the one-time setup to light it up.\n1. sign up at [posthog.com](https://posthog.com), pick **cloud-us**\n2. create one project: `torus`\n3. copy the **project api key** (`phc_xxxxxxxxxxxxx`) from project settings docs/posthog-setup.md https://github.com/torus-intelligence/torus/blob/staging/docs/posthog-setup.md github runbook active docs runbook docs" }, { "id": "doc:docs/superpowers/specs/2026-04-23-production-bootstrap-design.md", "kind": "doc", "title": "Production Bootstrap", "summary": "Bootstrap the Railway production environment to match staging infrastructure, fix CI failures on PR #110, and enable merge to main.\nPR #110 merges the full staging codebase to main for the first time. Production has:\n- 14 fossil DB tables (no migrations, no pgvector, no pg_trgm)\n- No S3 bucket", "path": "docs/superpowers/specs/2026-04-23-production-bootstrap-design.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/2026-04-23-production-bootstrap-design.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "production bootstrap bootstrap the railway production environment to match staging infrastructure, fix ci failures on pr #110, and enable merge to main.\npr #110 merges the full staging codebase to main for the first time. production has:\n- 14 fossil db tables (no migrations, no pgvector, no pg_trgm)\n- no s3 bucket docs/superpowers/specs/2026-04-23-production-bootstrap-design.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/2026-04-23-production-bootstrap-design.md github runbook active docs runbook docs superpowers specs" }, { "id": "skill:project-quarterly-review", "kind": "autopilot-skill", "title": "Project Quarterly Review", "summary": "---\nname: project-quarterly-review\ndescription: |\nOn-demand. Generates a quarterly readout for an EPC client.", "path": "ops/autopilot/skills/project-quarterly-review/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/project-quarterly-review/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "project-quarterly-review" ], "searchText": "project quarterly review ---\nname: project-quarterly-review\ndescription: |\non-demand. generates a quarterly readout for an epc client. ops/autopilot/skills/project-quarterly-review/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/project-quarterly-review/skill.md autopilot active autopilot skill skills project-quarterly-review" }, { "id": "skill:project-room", "kind": "autopilot-skill", "title": "Project Room", "summary": "---\nname: project-room\ndescription: |\nUse to maintain a per-project living brief. Append-only history of a Torus deployment for an EPC client.", "path": "ops/autopilot/skills/project-room/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/project-room/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "project-room" ], "searchText": "project room ---\nname: project-room\ndescription: |\nuse to maintain a per-project living brief. append-only history of a torus deployment for an epc client. ops/autopilot/skills/project-room/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/project-room/skill.md autopilot active autopilot skill skills project-room" }, { "id": "integration:railway", "kind": "integration", "title": "Railway Deploy Status", "summary": "Ops should expose deployment health, domains, service ownership, and stale volume cleanup state.", "source": "RAILWAY", "status": "planned", "tags": [ "railway", "deployments", "health" ], "nextAction": "Add read-only status cards through Railway API/MCP with redacted variables.", "searchText": "railway deploy status ops should expose deployment health, domains, service ownership, and stale volume cleanup state. railway planned add read-only status cards through railway api/mcp with redacted variables. railway deployments health" }, { "id": "autopilot-output:ops/autopilot/inventory/epc-sourcing/README.md", "kind": "autopilot-output", "title": "README", "summary": "Pull date: 2026-05-15\n- firms: 15238\n- projects: 2939\n- people: 0", "path": "ops/autopilot/inventory/epc-sourcing/README.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/epc-sourcing/README.md", "source": "AUTOPILOT", "category": "EPC_SOURCING", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "epc-sourcing" ], "nextAction": "Review sourced account and link to Company, Person, or Opportunity.", "searchText": "readme pull date: 2026-05-15\n- firms: 15238\n- projects: 2939\n- people: 0 ops/autopilot/inventory/epc-sourcing/readme.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/epc-sourcing/readme.md autopilot epc_sourcing needs-review review sourced account and link to company, person, or opportunity. autopilot review inventory epc-sourcing" }, { "id": "autopilot-output:ops/autopilot/inventory/knowledge/README.md", "kind": "autopilot-output", "title": "README", "summary": "Accumulated learnings and signals from Torus agents and team members.\nEvery knowledge entry uses this frontmatter:\nDescribe the finding. Include:\n- **Context:** what was happening when this was discovered", "path": "ops/autopilot/inventory/knowledge/README.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/README.md", "source": "AUTOPILOT", "category": "RESEARCH", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "knowledge" ], "nextAction": "Promote useful knowledge into runbooks, CRM notes, or Autopilot learnings.", "searchText": "readme accumulated learnings and signals from torus agents and team members.\nevery knowledge entry uses this frontmatter:\ndescribe the finding. include:\n- **context:** what was happening when this was discovered ops/autopilot/inventory/knowledge/readme.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/readme.md autopilot research needs-review promote useful knowledge into runbooks, crm notes, or autopilot learnings. autopilot review inventory knowledge" }, { "id": "doc:docs/superpowers/plans/2026-04-22-recurring-audit.md", "kind": "doc", "title": "Recurring Audit of Legend — Implementation Plan", "summary": "**Status:** open\n> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Run the Torus mega-audit in either `weekly` (full sweep) or `nightly` (diff-scoped with blast-radius fan-out) mode, producing a mega-report, Tier-1 auto-fix branch, and (weekly only) a master overhaul plan. All paths, branches, and output directories are date-stamped and derived at runtime.\n**Architecture:** Orchestrator (main session) resolves mode + date + diff, then dispatches pod agents in phases. Pods 1-6 in parallel (isolated worktrees, read-only). Pod 7 adversarial after. Merger. Master plan (weekly only). Tier-1 executor on writable wo...", "path": "docs/superpowers/plans/2026-04-22-recurring-audit.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-04-22-recurring-audit.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "recurring audit of legend — implementation plan **status:** open\n> **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** run the torus mega-audit in either `weekly` (full sweep) or `nightly` (diff-scoped with blast-radius fan-out) mode, producing a mega-report, tier-1 auto-fix branch, and (weekly only) a master overhaul plan. all paths, branches, and output directories are date-stamped and derived at runtime.\n**architecture:** orchestrator (main session) resolves mode + date + diff, then dispatches pod agents in phases. pods 1-6 in parallel (isolated worktrees, read-only). pod 7 adversarial after. merger. master plan (weekly only). tier-1 executor on writable wo... docs/superpowers/plans/2026-04-22-recurring-audit.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-04-22-recurring-audit.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/retrieval-strategy.md", "kind": "doc", "title": "Retrieval Strategy", "summary": "Entry point: `hybridSearch()` in `backend/plugins/engineering/search.ts`.\nAll agent context, workspace generation, and knowledge tools go through this function. Raw `SELECT FROM parameters` is only for CRUD and listing UI.\n**BM25** -- tsvector search with `ts_rank`, OR operator joins for broad matching. Searches parameters, graph_nodes, and document_chunks tables.\n**Vector** -- pgvector cosine similarity (`1 - (embedding <=> query_embedding)`) over graph_nodes and document_chunks. Embeddings are 1536-dim (OpenAI text-embedding-3-small) or 384-dim (sidecar bge-small-en-v1.5).", "path": "docs/retrieval-strategy.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/retrieval-strategy.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "retrieval strategy entry point: `hybridsearch()` in `backend/plugins/engineering/search.ts`.\nall agent context, workspace generation, and knowledge tools go through this function. raw `select from parameters` is only for crud and listing ui.\n**bm25** -- tsvector search with `ts_rank`, or operator joins for broad matching. searches parameters, graph_nodes, and document_chunks tables.\n**vector** -- pgvector cosine similarity (`1 - (embedding <=> query_embedding)`) over graph_nodes and document_chunks. embeddings are 1536-dim (openai text-embedding-3-small) or 384-dim (sidecar bge-small-en-v1.5). docs/retrieval-strategy.md https://github.com/torus-intelligence/torus/blob/staging/docs/retrieval-strategy.md github runbook active docs runbook docs" }, { "id": "skill:rfp-responder", "kind": "autopilot-skill", "title": "RFP/RFI Response Automator", "summary": "---\nname: rfp-responder\ndescription: |\nUse when responding to an RFP, RFI, or security questionnaire.", "path": "ops/autopilot/skills/rfp-responder/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/rfp-responder/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "rfp-responder" ], "searchText": "rfp/rfi response automator ---\nname: rfp-responder\ndescription: |\nuse when responding to an rfp, rfi, or security questionnaire. ops/autopilot/skills/rfp-responder/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/rfp-responder/skill.md autopilot active autopilot skill skills rfp-responder" }, { "id": "doc:ops/autopilot/context/sales-messaging.md", "kind": "doc", "title": "Sales Messaging -- Canonical Lines", "summary": "Engineering intelligence for EPC. Reads documents. Catches cross-discipline inconsistencies.\nTorus ingests engineering documents and flags inconsistencies across disciplines, codes, and document versions in minutes.\n1. **Cross-discipline review is manual.** Process / Mechanical / I&C / Civil / Structural deliverables don't agree, and the catches happen during commissioning.\n2. **Codes coverage is opaque.** ASME / API / NFPA references in design basis don't propagate visibly into every drawing.", "path": "ops/autopilot/context/sales-messaging.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/sales-messaging.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "sales messaging -- canonical lines engineering intelligence for epc. reads documents. catches cross-discipline inconsistencies.\ntorus ingests engineering documents and flags inconsistencies across disciplines, codes, and document versions in minutes.\n1. **cross-discipline review is manual.** process / mechanical / i&c / civil / structural deliverables don't agree, and the catches happen during commissioning.\n2. **codes coverage is opaque.** asme / api / nfpa references in design basis don't propagate visibly into every drawing. ops/autopilot/context/sales-messaging.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/sales-messaging.md autopilot autopilot active docs autopilot context" }, { "id": "autopilot-output:ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md", "kind": "autopilot-output", "title": "Scheduled Agent Mcp Bug", "summary": "---\ntitle: Scheduled agent MCP connectors silently fail to load\ncategory: technical\ndiscovered: 2026-04-01", "path": "ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md", "source": "AUTOPILOT", "category": "RESEARCH", "status": "needs-review", "tags": [ "autopilot", "review", "inventory", "knowledge" ], "nextAction": "Promote useful knowledge into runbooks, CRM notes, or Autopilot learnings.", "searchText": "scheduled agent mcp bug ---\ntitle: scheduled agent mcp connectors silently fail to load\ncategory: technical\ndiscovered: 2026-04-01 ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md autopilot research needs-review promote useful knowledge into runbooks, crm notes, or autopilot learnings. autopilot review inventory knowledge" }, { "id": "doc:ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md", "kind": "doc", "title": "Scheduled Agent MCP Bug", "summary": "---\ntitle: Scheduled agent MCP connectors silently fail to load\ncategory: technical\ndiscovered: 2026-04-01", "path": "ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "inventory", "knowledge" ], "searchText": "scheduled agent mcp bug ---\ntitle: scheduled agent mcp connectors silently fail to load\ncategory: technical\ndiscovered: 2026-04-01 ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/inventory/knowledge/scheduled-agent-mcp-bug.md autopilot autopilot active docs autopilot inventory knowledge" }, { "id": "doc:docs/seed-data-sources.md", "kind": "doc", "title": "Seed Data Sources", "summary": "Tier 1 reference data for Torus engineering intelligence. These are well-structured, well-labeled sources that should be ingested to power cross-reference consistency checks, parameter extraction, and code validation.\n`data/standards/CFIHOS-V2.0-Reference-Data-Library.xlsx`\n**Format:** Excel (11,891 master objects)\n**Status:** Seeded into Postgres (832 equipment classes, 1,388 properties, 4,707 mappings, 329 document types, 419 units, 305 standards)", "path": "docs/seed-data-sources.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/seed-data-sources.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "seed data sources tier 1 reference data for torus engineering intelligence. these are well-structured, well-labeled sources that should be ingested to power cross-reference consistency checks, parameter extraction, and code validation.\n`data/standards/cfihos-v2.0-reference-data-library.xlsx`\n**format:** excel (11,891 master objects)\n**status:** seeded into postgres (832 equipment classes, 1,388 properties, 4,707 mappings, 329 document types, 419 units, 305 standards) docs/seed-data-sources.md https://github.com/torus-intelligence/torus/blob/staging/docs/seed-data-sources.md github runbook active docs runbook docs" }, { "id": "doc:docs/skill-factory.md", "kind": "doc", "title": "Skill Factory", "summary": "Client-facing process automation. Observes how a firm handles RFIs, submittals, procurement approvals, change orders, etc. Learns the pattern. Eventually generates output.\nNot the same as Torus Autopilot (internal ops at `ops/autopilot/`). The Skill Factory lives at `backend/plugins/skills/`.\nVocabulary is the canonical L/G/C phases used across every substrate (see\n`backend/lifecycle/graduation-engine.ts`):", "path": "docs/skill-factory.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/skill-factory.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "skill factory client-facing process automation. observes how a firm handles rfis, submittals, procurement approvals, change orders, etc. learns the pattern. eventually generates output.\nnot the same as torus autopilot (internal ops at `ops/autopilot/`). the skill factory lives at `backend/plugins/skills/`.\nvocabulary is the canonical l/g/c phases used across every substrate (see\n`backend/lifecycle/graduation-engine.ts`): docs/skill-factory.md https://github.com/torus-intelligence/torus/blob/staging/docs/skill-factory.md github runbook active docs runbook docs" }, { "id": "doc:docs/architecture/skill-factory-boundary.md", "kind": "doc", "title": "Skill factory boundary", "summary": "Where \"skill\", \"process\", and \"playbook\" mean different things in this\ncodebase, and which is which.\nThere are **three** places where you'll see this kind of vocabulary. They\ndo different things, target different users, and have separate", "path": "docs/architecture/skill-factory-boundary.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/architecture/skill-factory-boundary.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "architecture" ], "searchText": "skill factory boundary where \"skill\", \"process\", and \"playbook\" mean different things in this\ncodebase, and which is which.\nthere are **three** places where you'll see this kind of vocabulary. they\ndo different things, target different users, and have separate docs/architecture/skill-factory-boundary.md https://github.com/torus-intelligence/torus/blob/staging/docs/architecture/skill-factory-boundary.md github architecture active docs architecture docs architecture" }, { "id": "doc:docs/superpowers/specs/INDEX.md", "kind": "doc", "title": "Specs Index", "summary": "One-line summary of every spec in `docs/superpowers/specs/`.\n- [2026-04-21-mega-audit-design.md](./2026-04-21-mega-audit-design.md) — Design spec for the overnight mega-audit: one mega-report + master overhaul plan + Tier-1 commit branch, ~3-4 hours overnight.\n- [2026-04-23-production-bootstrap-design.md](./2026-04-23-production-bootstrap-design.md) — Bootstrap the Railway production environment to match staging infrastructure, fix CI failures on PR #110, and enable merge to main.\n- [2026-05-05-mike-superset-absorption-roadmap.md](./2026-05-05-mike-superset-absorption-roadmap.md) — Cross-cutting meta-roadmap ranking 9 work buckets extracted from willchen96/mike and superset-sh/superset by dependency order.", "path": "docs/superpowers/specs/INDEX.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/specs/INDEX.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "specs" ], "searchText": "specs index one-line summary of every spec in `docs/superpowers/specs/`.\n- [2026-04-21-mega-audit-design.md](./2026-04-21-mega-audit-design.md) — design spec for the overnight mega-audit: one mega-report + master overhaul plan + tier-1 commit branch, ~3-4 hours overnight.\n- [2026-04-23-production-bootstrap-design.md](./2026-04-23-production-bootstrap-design.md) — bootstrap the railway production environment to match staging infrastructure, fix ci failures on pr #110, and enable merge to main.\n- [2026-05-05-mike-superset-absorption-roadmap.md](./2026-05-05-mike-superset-absorption-roadmap.md) — cross-cutting meta-roadmap ranking 9 work buckets extracted from willchen96/mike and superset-sh/superset by dependency order. docs/superpowers/specs/index.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/specs/index.md github runbook active docs runbook docs superpowers specs" }, { "id": "doc:docs/superpowers/plans/2026-05-21-staging-bugs-follow-up.md", "kind": "doc", "title": "Staging Bugs Follow-Up Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Fix four bugs admin@usetorus.com hit on staging after PR #45 deployed: Documents tab blank, Activity tab blank, Emails not clickable, People graph rendering hundreds of bogus \"orgs\" with no zoom.\n**Architecture:** Surgical per-symptom fixes. Three are pure frontend, one needs a new backend endpoint, one is a backend WHERE-clause tightening. No new dependencies.\n**Tech Stack:** Bun, TypeScript, SvelteKit (Svelte 5 runes), `bun:test`, Postgres 18.", "path": "docs/superpowers/plans/2026-05-21-staging-bugs-follow-up.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-staging-bugs-follow-up.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "staging bugs follow-up plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** fix four bugs admin@usetorus.com hit on staging after pr #45 deployed: documents tab blank, activity tab blank, emails not clickable, people graph rendering hundreds of bogus \"orgs\" with no zoom.\n**architecture:** surgical per-symptom fixes. three are pure frontend, one needs a new backend endpoint, one is a backend where-clause tightening. no new dependencies.\n**tech stack:** bun, typescript, sveltekit (svelte 5 runes), `bun:test`, postgres 18. docs/superpowers/plans/2026-05-21-staging-bugs-follow-up.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-staging-bugs-follow-up.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/standards-integration-plan.md", "kind": "doc", "title": "Standards Integration Plan: CFIHOS + DEXPI", "summary": "Industry-standard taxonomy for capital facilities. 11,891 master objects.\n| Dataset | Rows | Use |\n|---------|------|-----|\n| Equipment classes | 832 | Canonical equipment types (replace our ad-hoc types) |", "path": "docs/standards-integration-plan.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/standards-integration-plan.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs" ], "searchText": "standards integration plan: cfihos + dexpi industry-standard taxonomy for capital facilities. 11,891 master objects.\n| dataset | rows | use |\n|---------|------|-----|\n| equipment classes | 832 | canonical equipment types (replace our ad-hoc types) | docs/standards-integration-plan.md https://github.com/torus-intelligence/torus/blob/staging/docs/standards-integration-plan.md github runbook active docs runbook docs" }, { "id": "runner:standards-watch", "kind": "autopilot-runner", "title": "Standards Watch", "summary": "Scheduled or operator-run Autopilot workflow at ops/autopilot/runners/standards-watch/run.ts.", "path": "ops/autopilot/runners/standards-watch/run.ts", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/runners/standards-watch/run.ts", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "runner", "standards-watch" ], "nextAction": "Inspect runner state, latest output, and linked CRM artifacts.", "searchText": "standards watch scheduled or operator-run autopilot workflow at ops/autopilot/runners/standards-watch/run.ts. ops/autopilot/runners/standards-watch/run.ts https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/runners/standards-watch/run.ts autopilot active inspect runner state, latest output, and linked crm artifacts. autopilot runner standards-watch" }, { "id": "skill:standards-watch", "kind": "autopilot-skill", "title": "Standards Watch", "summary": "---\nname: standards-watch\ndescription: Scheduled RSS/API ingest of code-revision, CSB, EPA, project-award, and standards-deadline events into the event bank.\n---", "path": "ops/autopilot/skills/standards-watch/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/standards-watch/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "standards-watch" ], "searchText": "standards watch ---\nname: standards-watch\ndescription: scheduled rss/api ingest of code-revision, csb, epa, project-award, and standards-deadline events into the event bank.\n--- ops/autopilot/skills/standards-watch/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/standards-watch/skill.md autopilot active autopilot skill skills standards-watch" }, { "id": "doc:ops/autopilot/context/org/team.md", "kind": "doc", "title": "Team Roster", "summary": "| Name | Role | Discipline | Email | Discord | Linear | Joined |\n|------|------|------------|-------|---------|--------|--------|\n| Rahul Thayil | Founder / CEO | -- | rahul@usetorus.com | @rahul | @rahul | -- |\n(Fill in remaining team members as they join.)", "path": "ops/autopilot/context/org/team.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/org/team.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context", "org" ], "searchText": "team roster | name | role | discipline | email | discord | linear | joined |\n|------|------|------------|-------|---------|--------|--------|\n| rahul thayil | founder / ceo | -- | rahul@usetorus.com | @rahul | @rahul | -- |\n(fill in remaining team members as they join.) ops/autopilot/context/org/team.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/org/team.md autopilot autopilot active docs autopilot context org" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly/tier-1-execution-log.md", "kind": "doc", "title": "Tier-1 Execution Log -- 2026-05-21 Weekly Audit", "summary": "**Branch:** `audit-weekly-2026-05-21`\n**Base:** `spangle-flame` @ `8dcaa3d6` (staging tip + this run's 4 docs commits)\n**Executor:** Tier-1 auto-executor (Opus 4.7, 1M context)\n**Date:** 2026-05-21", "path": "docs/superpowers/audits/2026-05-21-weekly/tier-1-execution-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/tier-1-execution-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-21-weekly" ], "searchText": "tier-1 execution log -- 2026-05-21 weekly audit **branch:** `audit-weekly-2026-05-21`\n**base:** `spangle-flame` @ `8dcaa3d6` (staging tip + this run's 4 docs commits)\n**executor:** tier-1 auto-executor (opus 4.7, 1m context)\n**date:** 2026-05-21 docs/superpowers/audits/2026-05-21-weekly/tier-1-execution-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly/tier-1-execution-log.md github runbook active docs runbook docs superpowers audits 2026-05-21-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/tier-1-execution-log.md", "kind": "doc", "title": "Tier-1 Execution Log -- Weekly Audit 2026-05-20", "summary": "- **Branch:** `audit-weekly-2026-05-20` (off `pleasant-list`)\n- **Mode:** weekly\n- **v2 Tier-1 findings:** 47\n- **Concerns executed:** 9 commits", "path": "docs/superpowers/audits/2026-05-20-weekly/tier-1-execution-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/tier-1-execution-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "tier-1 execution log -- weekly audit 2026-05-20 - **branch:** `audit-weekly-2026-05-20` (off `pleasant-list`)\n- **mode:** weekly\n- **v2 tier-1 findings:** 47\n- **concerns executed:** 9 commits docs/superpowers/audits/2026-05-20-weekly/tier-1-execution-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/tier-1-execution-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:ops/autopilot/context/org/glossary.md", "kind": "doc", "title": "Torus / EPC Glossary", "summary": "| Term | Definition |\n|------|-----------|\n| CFIHOS | Capital Facilities Information Handover Specification (V2.0). Standard for engineering data handover. |\n| FEL1 / FEL2 / FEL3 | Front-End Loading phases. FEL1 = feasibility, FEL2 = concept, FEL3 = detailed design. Torus sweet spot is FEL2/FEL3. |", "path": "ops/autopilot/context/org/glossary.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/org/glossary.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context", "org" ], "searchText": "torus / epc glossary | term | definition |\n|------|-----------|\n| cfihos | capital facilities information handover specification (v2.0). standard for engineering data handover. |\n| fel1 / fel2 / fel3 | front-end loading phases. fel1 = feasibility, fel2 = concept, fel3 = detailed design. torus sweet spot is fel2/fel3. | ops/autopilot/context/org/glossary.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/org/glossary.md autopilot autopilot active docs autopilot context org" }, { "id": "doc:docs/architecture.md", "kind": "doc", "title": "Torus Architecture", "summary": "Torus is a backend service with a plugin system, a Python ML sidecar, a SvelteKit frontend, and an Astro marketing site. Cloud deploys to Railway. Sovereign clients get Docker Compose with self-hosted inference.\nThe backend (`backend/`) is a Bun HTTP server with Postgres. Core routes handle hooks (webhook inbox), stored views (read-only queries), health, and auth guard plumbing. Product capabilities live in plugins.\nPlugins live in `backend/plugins//`. Each exports a `register(app)` function that adds routes at `/plugins//`.\n| Plugin | Routes | Purpose |", "path": "docs/architecture.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/architecture.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs" ], "searchText": "torus architecture torus is a backend service with a plugin system, a python ml sidecar, a sveltekit frontend, and an astro marketing site. cloud deploys to railway. sovereign clients get docker compose with self-hosted inference.\nthe backend (`backend/`) is a bun http server with postgres. core routes handle hooks (webhook inbox), stored views (read-only queries), health, and auth guard plumbing. product capabilities live in plugins.\nplugins live in `backend/plugins//`. each exports a `register(app)` function that adds routes at `/plugins//`.\n| plugin | routes | purpose | docs/architecture.md https://github.com/torus-intelligence/torus/blob/staging/docs/architecture.md github architecture active docs architecture docs" }, { "id": "doc:ops/autopilot/context/brand.md", "kind": "doc", "title": "Torus Brand Identity", "summary": "| Name | Hex | Usage |\n|------|-----|-------|\n| Black | #000000 | Backgrounds, dark theme |\n| Gray 950 | #080808 | Deep background |", "path": "ops/autopilot/context/brand.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/brand.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "torus brand identity | name | hex | usage |\n|------|-----|-------|\n| black | #000000 | backgrounds, dark theme |\n| gray 950 | #080808 | deep background | ops/autopilot/context/brand.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/brand.md autopilot autopilot active docs autopilot context" }, { "id": "doc:ops/crm/README.md", "kind": "doc", "title": "Torus CRM", "summary": "This module runs [Twenty](https://github.com/twentyhq/twenty) as Torus's internal CRM.\nThe deployment intentionally uses the upstream `twentycrm/twenty` image instead of vendoring the Twenty monorepo. Keep this as an ops wrapper unless we need source-level product changes.\nUse a dedicated Railway module with separate data stores:\n| Railway service | Root dir | Purpose | Public |", "path": "ops/crm/README.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/crm/README.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "crm" ], "searchText": "torus crm this module runs [twenty](https://github.com/twentyhq/twenty) as torus's internal crm.\nthe deployment intentionally uses the upstream `twentycrm/twenty` image instead of vendoring the twenty monorepo. keep this as an ops wrapper unless we need source-level product changes.\nuse a dedicated railway module with separate data stores:\n| railway service | root dir | purpose | public | ops/crm/readme.md https://github.com/torus-intelligence/torus/blob/staging/ops/crm/readme.md github runbook active docs runbook crm" }, { "id": "skill:torus-report", "kind": "autopilot-skill", "title": "Torus Report", "summary": "---\nname: torus-report\ndescription: |\nUse to generate a one-shot publication-ready HTML report (gauge or line chart).", "path": "ops/autopilot/skills/torus-report/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/torus-report/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "torus-report" ], "searchText": "torus report ---\nname: torus-report\ndescription: |\nuse to generate a one-shot publication-ready html report (gauge or line chart). ops/autopilot/skills/torus-report/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/torus-report/skill.md autopilot active autopilot skill skills torus-report" }, { "id": "doc:ops/autopilot/context/repos.md", "kind": "doc", "title": "Torus Repositories", "summary": "| Repo | Description | Primary Language |\n|------|------------|-----------------|\n| torus | Monorepo -- backend, frontend, website, autopilot | TypeScript |\nAll repos are under `https://github.com/Torus-Intelligence/`. Clone via:", "path": "ops/autopilot/context/repos.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/context/repos.md", "source": "AUTOPILOT", "category": "AUTOPILOT", "status": "active", "tags": [ "docs", "autopilot", "context" ], "searchText": "torus repositories | repo | description | primary language |\n|------|------------|-----------------|\n| torus | monorepo -- backend, frontend, website, autopilot | typescript |\nall repos are under `https://github.com/torus-intelligence/`. clone via: ops/autopilot/context/repos.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/context/repos.md autopilot autopilot active docs autopilot context" }, { "id": "skill:torus-voice", "kind": "autopilot-skill", "title": "Torus Voice Skill", "summary": "---\nname: torus-voice\ndescription: |\nUse when drafting any communication as Rahul Thayil -- emails, Discord messages,", "path": "ops/autopilot/skills/torus-voice/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/torus-voice/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "torus-voice" ], "searchText": "torus voice skill ---\nname: torus-voice\ndescription: |\nuse when drafting any communication as rahul thayil -- emails, discord messages, ops/autopilot/skills/torus-voice/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/torus-voice/skill.md autopilot active autopilot skill skills torus-voice" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly.md", "kind": "doc", "title": "Torus weekly Audit — 2026-05-20 (v2 deep)", "summary": "This is the v2 (deep) pass of the weekly mega-audit, run 2026-05-20 against the orchestrator worktree at `/Users/rahulthayil/.superset/worktrees/torus/pleasant-list`. Where v1 produced 131 findings, v2 produced 306 raw findings across the seven pods; after dedup across file:line + problem identity, 246 distinct findings remain. Pod 7 (adversarial) ran two rounds of stress-testing and gap-discovery on top of pods 1-6 and is the source of the largest single corrections.\nThe three things that matter most from this run:\n1. **The login bypass is the same shape as v1 but the file:line and reasoning are different.** v1 #1 claimed Bun.password / bcrypt did not exist; pod-7 v1 rejected that and v2 confirms — bcrypt is wired correctly (`backend/plugins/auth/passwords.ts:7`, `backend/plugins/auth/in...", "path": "docs/superpowers/audits/2026-05-20-weekly.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits" ], "searchText": "torus weekly audit — 2026-05-20 (v2 deep) this is the v2 (deep) pass of the weekly mega-audit, run 2026-05-20 against the orchestrator worktree at `/users/rahulthayil/.superset/worktrees/torus/pleasant-list`. where v1 produced 131 findings, v2 produced 306 raw findings across the seven pods; after dedup across file:line + problem identity, 246 distinct findings remain. pod 7 (adversarial) ran two rounds of stress-testing and gap-discovery on top of pods 1-6 and is the source of the largest single corrections.\nthe three things that matter most from this run:\n1. **the login bypass is the same shape as v1 but the file:line and reasoning are different.** v1 #1 claimed bun.password / bcrypt did not exist; pod-7 v1 rejected that and v2 confirms — bcrypt is wired correctly (`backend/plugins/auth/passwords.ts:7`, `backend/plugins/auth/in... docs/superpowers/audits/2026-05-20-weekly.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly.md github runbook active docs runbook docs superpowers audits" }, { "id": "doc:docs/superpowers/audits/2026-05-21-weekly.md", "kind": "doc", "title": "Torus Weekly Audit — 2026-05-21 (save-point)", "summary": "**Save-point status:** PR #44 (merged 2026-05-20) landed Wave 0-4 cleanup against the v2 mega-audit. Independent re-verification across 7 pods at staging tip `8125d2f1` confirms roughly 75% of last week's findings are closed at the file:line cited, with the residual concentrated in (a) catch-all sub-routes that escaped grep-based middleware sweeps, (b) infrastructure that landed without adoption (`withIdempotency`, `parameters_trusted`, `graph_all_nodes_mv_rls`), and (c) durability primitives whose production code is correct but whose tests are `test.skip` placeholders.\n**Three load-bearing items this cycle:**\n1. **Pod 7 found two new CRITICALs that prior pods missed.** FINDING-pod-7-1: the canonical `graph_all_nodes` view's graph_nodes leg omits the trust filter — unapproved generated do...", "path": "docs/superpowers/audits/2026-05-21-weekly.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits" ], "searchText": "torus weekly audit — 2026-05-21 (save-point) **save-point status:** pr #44 (merged 2026-05-20) landed wave 0-4 cleanup against the v2 mega-audit. independent re-verification across 7 pods at staging tip `8125d2f1` confirms roughly 75% of last week's findings are closed at the file:line cited, with the residual concentrated in (a) catch-all sub-routes that escaped grep-based middleware sweeps, (b) infrastructure that landed without adoption (`withidempotency`, `parameters_trusted`, `graph_all_nodes_mv_rls`), and (c) durability primitives whose production code is correct but whose tests are `test.skip` placeholders.\n**three load-bearing items this cycle:**\n1. **pod 7 found two new criticals that prior pods missed.** finding-pod-7-1: the canonical `graph_all_nodes` view's graph_nodes leg omits the trust filter — unapproved generated do... docs/superpowers/audits/2026-05-21-weekly.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-21-weekly.md github runbook active docs runbook docs superpowers audits" }, { "id": "doc:docs/superpowers/plans/2026-05-20-weekly-overhaul.md", "kind": "doc", "title": "Torus Weekly Overhaul Plan — 2026-05-20 (v2 deep)", "summary": "**Status:** open\n**Source mega-report:** `docs/superpowers/audits/2026-05-20-weekly.md` (246 findings post-dedup)\n**Source pods:** `docs/superpowers/audits/2026-05-20-weekly/pod-{1..7}-*.md`\n**Design spec:** `docs/superpowers/specs/2026-04-21-mega-audit-design.md`", "path": "docs/superpowers/plans/2026-05-20-weekly-overhaul.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-20-weekly-overhaul.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "torus weekly overhaul plan — 2026-05-20 (v2 deep) **status:** open\n**source mega-report:** `docs/superpowers/audits/2026-05-20-weekly.md` (246 findings post-dedup)\n**source pods:** `docs/superpowers/audits/2026-05-20-weekly/pod-{1..7}-*.md`\n**design spec:** `docs/superpowers/specs/2026-04-21-mega-audit-design.md` docs/superpowers/plans/2026-05-20-weekly-overhaul.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-20-weekly-overhaul.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/superpowers/plans/2026-05-21-weekly-overhaul.md", "kind": "doc", "title": "Torus Weekly Overhaul Plan — 2026-05-21", "summary": "**Status:** open\n**Source mega-report:** `docs/superpowers/audits/2026-05-21-weekly.md` (116 distinct findings post-dedup)\n**Source pods:** `docs/superpowers/audits/2026-05-21-weekly/pod-{1..7}-*.md`\n**Design spec:** `docs/superpowers/specs/2026-04-21-mega-audit-design.md`", "path": "docs/superpowers/plans/2026-05-21-weekly-overhaul.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-weekly-overhaul.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "torus weekly overhaul plan — 2026-05-21 **status:** open\n**source mega-report:** `docs/superpowers/audits/2026-05-21-weekly.md` (116 distinct findings post-dedup)\n**source pods:** `docs/superpowers/audits/2026-05-21-weekly/pod-{1..7}-*.md`\n**design spec:** `docs/superpowers/specs/2026-04-21-mega-audit-design.md` docs/superpowers/plans/2026-05-21-weekly-overhaul.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-21-weekly-overhaul.md github runbook active docs runbook docs superpowers plans" }, { "id": "doc:docs/architecture/eval-systems.md", "kind": "doc", "title": "Two eval systems (and why they stay separate for now)", "summary": "Torus has two eval surfaces. They were written for different jobs and\nshare zero infrastructure today. The synthesis audit flagged this as a\nfollow-up cleanup; on closer inspection, deeper unification carries\nmore risk than payoff, so this document captures what they are, what's", "path": "docs/architecture/eval-systems.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/architecture/eval-systems.md", "source": "GITHUB", "category": "ARCHITECTURE", "status": "active", "tags": [ "docs", "architecture", "docs", "architecture" ], "searchText": "two eval systems (and why they stay separate for now) torus has two eval surfaces. they were written for different jobs and\nshare zero infrastructure today. the synthesis audit flagged this as a\nfollow-up cleanup; on closer inspection, deeper unification carries\nmore risk than payoff, so this document captures what they are, what's docs/architecture/eval-systems.md https://github.com/torus-intelligence/torus/blob/staging/docs/architecture/eval-systems.md github architecture active docs architecture docs architecture" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/unused-but-built-todos.md", "kind": "doc", "title": "Unused-but-Built — TODO review (do NOT auto-delete)", "summary": "Origin: 2026-05-20 weekly audit, Cluster 3b. The audit flagged these as \"dead code (0 callers)\". On manual review, every one is either explicitly part of an active plan or an in-progress feature that just hasn't been wired up yet. Axing them would delete in-flight product work.\n**Cluster 3b branch `audit-w1-cluster-3b` is DO-NOT-MERGE-AS-IS.** Cherry-pick only:\n- W1-57 (`backend/plugins/resolver-sdk/index.ts` stub) — safe addition.\n- W1-101 (`NodeGraph.svelte` header note) — safe addition.", "path": "docs/superpowers/audits/2026-05-20-weekly/unused-but-built-todos.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/unused-but-built-todos.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "unused-but-built — todo review (do not auto-delete) origin: 2026-05-20 weekly audit, cluster 3b. the audit flagged these as \"dead code (0 callers)\". on manual review, every one is either explicitly part of an active plan or an in-progress feature that just hasn't been wired up yet. axing them would delete in-flight product work.\n**cluster 3b branch `audit-w1-cluster-3b` is do-not-merge-as-is.** cherry-pick only:\n- w1-57 (`backend/plugins/resolver-sdk/index.ts` stub) — safe addition.\n- w1-101 (`nodegraph.svelte` header note) — safe addition. docs/superpowers/audits/2026-05-20-weekly/unused-but-built-todos.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/unused-but-built-todos.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-0-execution-log.md", "kind": "doc", "title": "Wave 0 Hotfix Execution Log -- 2026-05-20", "summary": "Tracks Wave 0 hotfix application against `audit-weekly-2026-05-20`. One row\nper concern. Entries are appended as they are committed.\n| Entry | Status | Notes |\n|-------|--------|-------|", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-0-execution-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-0-execution-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 0 hotfix execution log -- 2026-05-20 tracks wave 0 hotfix application against `audit-weekly-2026-05-20`. one row\nper concern. entries are appended as they are committed.\n| entry | status | notes |\n|-------|--------|-------| docs/superpowers/audits/2026-05-20-weekly/wave-0-execution-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-0-execution-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-1-log.md", "kind": "doc", "title": "Wave 1 Cluster 1 — Schema drift fixes", "summary": "Branch: `audit-w1-cluster-1` (based on `origin/audit-weekly-2026-05-20`)\nRun date: 2026-05-20\nSchema-drift entries from the weekly audit plan. Every fix anchors a code\nwrite or read against `db/schema.sql` ground truth, removes references to", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-1-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-1-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 1 cluster 1 — schema drift fixes branch: `audit-w1-cluster-1` (based on `origin/audit-weekly-2026-05-20`)\nrun date: 2026-05-20\nschema-drift entries from the weekly audit plan. every fix anchors a code\nwrite or read against `db/schema.sql` ground truth, removes references to docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-1-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-1-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-2-log.md", "kind": "doc", "title": "Wave 1 Cluster 2 Execution Log -- 2026-05-20", "summary": "Tracks Wave 1 (Fetch / SSE / agent runtime) hotfix application against\n`audit-w1-cluster-2`. One row per concern. Entries are appended as commits land.\n| Entry | Status | Notes |\n|-------|--------|-------|", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-2-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-2-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 1 cluster 2 execution log -- 2026-05-20 tracks wave 1 (fetch / sse / agent runtime) hotfix application against\n`audit-w1-cluster-2`. one row per concern. entries are appended as commits land.\n| entry | status | notes |\n|-------|--------|-------| docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-2-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-2-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-4-log.md", "kind": "doc", "title": "Wave 1 Cluster 4 Execution Log -- 2026-05-20", "summary": "Email + graph plugin scoped fixes against `audit-weekly-2026-05-20`. One\ncommit-tracked entry per concern.\n| Entry | Status | Notes |\n|-------|--------|-------|", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-4-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-4-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 1 cluster 4 execution log -- 2026-05-20 email + graph plugin scoped fixes against `audit-weekly-2026-05-20`. one\ncommit-tracked entry per concern.\n| entry | status | notes |\n|-------|--------|-------| docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-4-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-4-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-c-log.md", "kind": "doc", "title": "Wave 2 Cluster C Execution Log -- 2026-05-20", "summary": "Tracks Wave 2 (Network / streaming / async helpers) consolidation against\n`audit-w2-cluster-c`. One row per concern. Entries are appended as commits land.\n| Entry | Status | Notes |\n|-------|--------|-------|", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-c-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-c-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 2 cluster c execution log -- 2026-05-20 tracks wave 2 (network / streaming / async helpers) consolidation against\n`audit-w2-cluster-c`. one row per concern. entries are appended as commits land.\n| entry | status | notes |\n|-------|--------|-------| docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-c-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-c-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-e-log.md", "kind": "doc", "title": "Wave 2 Cluster E — PlanRunner durability (W2-17 retry + W2-18 watchdog)", "summary": "Branch: `audit-w2-cluster-e` (based on `origin/audit-weekly-2026-05-20`)\nRun date: 2026-05-20\nTwo safe, additive slices of PlanRunner durability. Tier-3 surface\n(`backend/plugins/agent/runner.ts`) -- careful boundary discipline.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-e-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-e-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 2 cluster e — planrunner durability (w2-17 retry + w2-18 watchdog) branch: `audit-w2-cluster-e` (based on `origin/audit-weekly-2026-05-20`)\nrun date: 2026-05-20\ntwo safe, additive slices of planrunner durability. tier-3 surface\n(`backend/plugins/agent/runner.ts`) -- careful boundary discipline. docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-e-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-e-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-a-log.md", "kind": "doc", "title": "Wave 3 Cluster A — Break Plugin Cycles", "summary": "Run date: 2026-05-20\nBranch: `audit-w3-cluster-a`\nBase: `audit-weekly-2026-05-20` @ 8f5ce7ca\nFINDING-pod-4-v2-15. Three confirmed cross-plugin static import cycles", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-a-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-a-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 3 cluster a — break plugin cycles run date: 2026-05-20\nbranch: `audit-w3-cluster-a`\nbase: `audit-weekly-2026-05-20` @ 8f5ce7ca\nfinding-pod-4-v2-15. three confirmed cross-plugin static import cycles docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-a-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-a-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-b-log.md", "kind": "doc", "title": "Wave 4 — Cluster B Execution Log", "summary": "**Branch:** `audit-w4-cluster-b`\n**Base:** `audit-weekly-2026-05-20` @ `66ad061d`\n**Run date:** 2026-05-20\n**Scope:** W4-3 (requireFirmAccess sweep) + W4-22 (SSE replay leak) + W4-42 (workspace /chat firm mismatch, folded into W4-3) + W4-49 (catch-all routes, folded into W4-3)", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-b-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-b-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 4 — cluster b execution log **branch:** `audit-w4-cluster-b`\n**base:** `audit-weekly-2026-05-20` @ `66ad061d`\n**run date:** 2026-05-20\n**scope:** w4-3 (requirefirmaccess sweep) + w4-22 (sse replay leak) + w4-42 (workspace /chat firm mismatch, folded into w4-3) + w4-49 (catch-all routes, folded into w4-3) docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-b-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-b-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-e-log.md", "kind": "doc", "title": "Wave 4 — Cluster E execution log", "summary": "Branch: `audit-w4-cluster-e` (base: `origin/audit-weekly-2026-05-20`).\nScope: PlanRunner state machine (W4-8 + W4-31) and stuck-plan watchdog verification (W4-45).\nDate: 2026-05-20.\n**Decision applied:** Option 1 (persist cancel intent on the row; remove", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-e-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-e-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 4 — cluster e execution log branch: `audit-w4-cluster-e` (base: `origin/audit-weekly-2026-05-20`).\nscope: planrunner state machine (w4-8 + w4-31) and stuck-plan watchdog verification (w4-45).\ndate: 2026-05-20.\n**decision applied:** option 1 (persist cancel intent on the row; remove docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-e-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-e-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-f-log.md", "kind": "doc", "title": "Wave 4 — Cluster F execution log", "summary": "Branch: `audit-w4-cluster-f` (off `audit-weekly-2026-05-20`)\nRUN_DATE: 2026-05-20\nPosture: **preserve data, no destructive deletions**\nEntries (sequential, smallest first):", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-f-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-f-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 4 — cluster f execution log branch: `audit-w4-cluster-f` (off `audit-weekly-2026-05-20`)\nrun_date: 2026-05-20\nposture: **preserve data, no destructive deletions**\nentries (sequential, smallest first): docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-f-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-f-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-i-log.md", "kind": "doc", "title": "Wave 4 — Cluster I execution log", "summary": "**Run date:** 2026-05-20\n**Branch:** `audit-w4-cluster-i` off `audit-weekly-2026-05-20`\n**Scope:** Misc remaining items — W4-2, W4-4, W4-9, W4-25 prep, W4-26 verify, W4-28, W4-36, W4-37, W4-39, W4-44, W4-47, W4-48.\n- Verified `allow_cross_backend_fallback: false` in", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-i-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-i-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 4 — cluster i execution log **run date:** 2026-05-20\n**branch:** `audit-w4-cluster-i` off `audit-weekly-2026-05-20`\n**scope:** misc remaining items — w4-2, w4-4, w4-9, w4-25 prep, w4-26 verify, w4-28, w4-36, w4-37, w4-39, w4-44, w4-47, w4-48.\n- verified `allow_cross_backend_fallback: false` in docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-i-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-i-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-decisions.md", "kind": "doc", "title": "Wave 4 — Decision map (from user 2026-05-20)", "summary": "Per-entry decision for each of the 49 Tier-3 items in the master plan. Items marked **DEFER → wave-4-todo.md** are not executed this run.\nBuild-mode posture from the user: **preserve data first**, handle GDPR / deletion / quota concerns later when the product matures.\n| Sub | Decision | Action |\n|---|---|---|", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-decisions.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-decisions.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 4 — decision map (from user 2026-05-20) per-entry decision for each of the 49 tier-3 items in the master plan. items marked **defer → wave-4-todo.md** are not executed this run.\nbuild-mode posture from the user: **preserve data first**, handle gdpr / deletion / quota concerns later when the product matures.\n| sub | decision | action |\n|---|---|---| docs/superpowers/audits/2026-05-20-weekly/wave-4-decisions.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-decisions.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-todo.md", "kind": "doc", "title": "Wave 4 — TODO (deferred decisions)", "summary": "Items the user explicitly chose to defer during the 2026-05-20 audit Wave-4 decision pass. Each one is a real risk that we're consciously NOT closing this run. Revisit when product reaches the relevant maturity.\n- **Why deferred:** Overkill right now. We don't have the user volume that makes session revocation a daily operation.\n- **What needs to ship eventually:** Server-side `iat`/`exp`/`jti` in the cookie payload; `sessions` table with revocation flag; invalidate-on-password-change.\n- **Trigger to revisit:** First incident where we can't kill an active session for a fired user / leaked device.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-todo.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-todo.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 4 — todo (deferred decisions) items the user explicitly chose to defer during the 2026-05-20 audit wave-4 decision pass. each one is a real risk that we're consciously not closing this run. revisit when product reaches the relevant maturity.\n- **why deferred:** overkill right now. we don't have the user volume that makes session revocation a daily operation.\n- **what needs to ship eventually:** server-side `iat`/`exp`/`jti` in the cookie payload; `sessions` table with revocation flag; invalidate-on-password-change.\n- **trigger to revisit:** first incident where we can't kill an active session for a fired user / leaked device. docs/superpowers/audits/2026-05-20-weekly/wave-4-todo.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-todo.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-5-code-review-log.md", "kind": "doc", "title": "Wave 5 — Code-review pass execution log (2026-05-21)", "summary": "Post-merge code review of PR #44 (`audit-weekly-2026-05-20`) surfaced 30+ residual findings across seven scopes: migrations, auth + authorization, RLS + trust filter, security headers + sandboxing, agent / plan runner, plugin cycles + workspace decomposition, Wave-2 helpers. All findings closed in two commits on the PR branch before merge:\n- `76269179` — code fixes\n- `919a606f` — CI greenup (storage mock leak, sidecar ruff, schema regen)\nThis log records what closed and what was deferred. Format follows the per-cluster pattern (`wave-N-cluster-X-log.md`).", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-5-code-review-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-5-code-review-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave 5 — code-review pass execution log (2026-05-21) post-merge code review of pr #44 (`audit-weekly-2026-05-20`) surfaced 30+ residual findings across seven scopes: migrations, auth + authorization, rls + trust filter, security headers + sandboxing, agent / plan runner, plugin cycles + workspace decomposition, wave-2 helpers. all findings closed in two commits on the pr branch before merge:\n- `76269179` — code fixes\n- `919a606f` — ci greenup (storage mock leak, sidecar ruff, schema regen)\nthis log records what closed and what was deferred. format follows the per-cluster pattern (`wave-n-cluster-x-log.md`). docs/superpowers/audits/2026-05-20-weekly/wave-5-code-review-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-5-code-review-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3a-log.md", "kind": "doc", "title": "Wave-1 Cluster 3a — Dep CVE patches + cross-workspace dedup", "summary": "**Run date:** 2026-05-20\n**Branch:** `audit-w1-cluster-3a`\n**Base:** `audit-weekly-2026-05-20`\n**Scope:** 5 dependency-hygiene findings from pod-1-v2 and pod-7-v2.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3a-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3a-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 3a — dep cve patches + cross-workspace dedup **run date:** 2026-05-20\n**branch:** `audit-w1-cluster-3a`\n**base:** `audit-weekly-2026-05-20`\n**scope:** 5 dependency-hygiene findings from pod-1-v2 and pod-7-v2. docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3a-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3a-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3b-log.md", "kind": "doc", "title": "Wave-1 Cluster 3b — Selective Cherry-Pick Log", "summary": "RUN_DATE: 2026-05-20\nBranch: audit-w1-cluster-3b (selective cherry-pick only)\nBase: origin/audit-weekly-2026-05-20\nCluster 3b on origin contained six commits. The orchestrator merge plan", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3b-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3b-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 3b — selective cherry-pick log run_date: 2026-05-20\nbranch: audit-w1-cluster-3b (selective cherry-pick only)\nbase: origin/audit-weekly-2026-05-20\ncluster 3b on origin contained six commits. the orchestrator merge plan docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3b-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-3b-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-5-log.md", "kind": "doc", "title": "Wave-1 Cluster 5 Execution Log — Perf (N+1, indexes, bundles)", "summary": "**Run date:** 2026-05-20\n**Branch:** audit-w1-cluster-5\n**Scope:** Tier-2 performance hot paths. Tier-3 surfaces deferred.\nConverted each per-row loop to a single UNNEST-based batch statement following", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-5-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-5-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 5 execution log — perf (n+1, indexes, bundles) **run date:** 2026-05-20\n**branch:** audit-w1-cluster-5\n**scope:** tier-2 performance hot paths. tier-3 surfaces deferred.\nconverted each per-row loop to a single unnest-based batch statement following docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-5-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-5-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-6-log.md", "kind": "doc", "title": "Wave-1 Cluster 6 -- Build, CI, scripts", "summary": "RUN_DATE: 2026-05-20\nBranch: audit-w1-cluster-6\nBase: origin/audit-weekly-2026-05-20\nBuild, CI, scripts. 7 entries.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-6-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-6-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 6 -- build, ci, scripts run_date: 2026-05-20\nbranch: audit-w1-cluster-6\nbase: origin/audit-weekly-2026-05-20\nbuild, ci, scripts. 7 entries. docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-6-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-6-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-7-log.md", "kind": "doc", "title": "Wave-1 Cluster 7 Execution Log -- Weekly Audit 2026-05-20", "summary": "- **Branch:** `audit-w1-cluster-7` (off `audit-weekly-2026-05-20`)\n- **Focus:** Frontend a11y + design tokens (pod-6 v2 follow-up)\n- **Entries assigned:** 14\n- **Concerns executed:** 7 commits", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-7-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-7-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 7 execution log -- weekly audit 2026-05-20 - **branch:** `audit-w1-cluster-7` (off `audit-weekly-2026-05-20`)\n- **focus:** frontend a11y + design tokens (pod-6 v2 follow-up)\n- **entries assigned:** 14\n- **concerns executed:** 7 commits docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-7-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-7-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-8-log.md", "kind": "doc", "title": "Wave-1 Cluster 8 -- Frontend SSR + page decomposition + transactional writes", "summary": "- **Branch:** `audit-w1-cluster-8` (off `audit-weekly-2026-05-20`)\n- **Worktree:** `/Users/rahulthayil/.superset/projects/torus/.claude/worktrees/agent-aa6ce3b2604a6961f`\n- **Run date:** 2026-05-20\n- **Commits:** 8", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-8-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-8-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 8 -- frontend ssr + page decomposition + transactional writes - **branch:** `audit-w1-cluster-8` (off `audit-weekly-2026-05-20`)\n- **worktree:** `/users/rahulthayil/.superset/projects/torus/.claude/worktrees/agent-aa6ce3b2604a6961f`\n- **run date:** 2026-05-20\n- **commits:** 8 docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-8-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-8-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-9-log.md", "kind": "doc", "title": "Wave-1 Cluster 9 -- Docs, license, sovereign, governance", "summary": "**Run date:** 2026-05-20\n**Branch:** audit-w1-cluster-9 (off origin/audit-weekly-2026-05-20)\nItems addressed below. Status convention:\n- DONE -- code or docs landed.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-9-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-9-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-1 cluster 9 -- docs, license, sovereign, governance **run date:** 2026-05-20\n**branch:** audit-w1-cluster-9 (off origin/audit-weekly-2026-05-20)\nitems addressed below. status convention:\n- done -- code or docs landed. docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-9-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-1-cluster-9-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-a-log.md", "kind": "doc", "title": "Wave-2 Cluster A -- HTTP middleware (auth, rate-limit, idempotency, errors, Zod)", "summary": "**Run date:** 2026-05-20\n**Branch:** audit-w2-cluster-a (off origin/audit-weekly-2026-05-20 @ 23a26f07)\nStatus convention:\n- DONE -- shipped this cluster.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-a-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-a-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-2 cluster a -- http middleware (auth, rate-limit, idempotency, errors, zod) **run date:** 2026-05-20\n**branch:** audit-w2-cluster-a (off origin/audit-weekly-2026-05-20 @ 23a26f07)\nstatus convention:\n- done -- shipped this cluster. docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-a-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-a-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-b-log.md", "kind": "doc", "title": "Wave-2 Cluster B -- DB / RLS / data helpers", "summary": "**Run date:** 2026-05-20\n**Branch:** audit-w2-cluster-b (off origin/audit-weekly-2026-05-20)\nStatus convention: DONE = code landed; DEFERRED = cross-cluster dependency\nnot yet shipped; LOGGED = called out for a follow-up sweep.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-b-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-b-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-2 cluster b -- db / rls / data helpers **run date:** 2026-05-20\n**branch:** audit-w2-cluster-b (off origin/audit-weekly-2026-05-20)\nstatus convention: done = code landed; deferred = cross-cluster dependency\nnot yet shipped; logged = called out for a follow-up sweep. docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-b-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-b-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-d-log.md", "kind": "doc", "title": "Wave-2 Cluster D execution log", "summary": "**Run date:** 2026-05-20\n**Branch:** `audit-w2-cluster-d` off `audit-weekly-2026-05-20`\n**Scope:** LLM gateway hardening — W2-7, W2-8, W2-16, W2-19, W2-25\nFive tightly-coupled changes to `backend/plugins/llm/` and the agent", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-d-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-d-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-2 cluster d execution log **run date:** 2026-05-20\n**branch:** `audit-w2-cluster-d` off `audit-weekly-2026-05-20`\n**scope:** llm gateway hardening — w2-7, w2-8, w2-16, w2-19, w2-25\nfive tightly-coupled changes to `backend/plugins/llm/` and the agent docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-d-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-d-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-f-log.md", "kind": "doc", "title": "Wave-2 Cluster F Execution Log — Storage + Sandbox Helpers", "summary": "**Run date:** 2026-05-20\n**Branch:** audit-w2-cluster-f\n**Scope:** W2-9 (storage lifecycle + sweeper extension), W2-24 (unified sandboxContent helper).\n**Lifecycle spec.** `deploy/storage-lifecycle.json` + `deploy/storage-lifecycle.README.md`.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-f-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-f-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-2 cluster f execution log — storage + sandbox helpers **run date:** 2026-05-20\n**branch:** audit-w2-cluster-f\n**scope:** w2-9 (storage lifecycle + sweeper extension), w2-24 (unified sandboxcontent helper).\n**lifecycle spec.** `deploy/storage-lifecycle.json` + `deploy/storage-lifecycle.readme.md`. docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-f-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-2-cluster-f-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-b-log.md", "kind": "doc", "title": "Wave-3 Cluster B -- Schema + migrations + KG cleanup", "summary": "**Run date:** 2026-05-20\n**Branch:** audit-w3-cluster-b (off origin/audit-weekly-2026-05-20)\nStatus convention: DONE = code landed; LOGGED = called out for a follow-up.\n`db/migrations/0088_fk_supporting_indexes.sql`. Audited every FK in", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-b-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-b-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-3 cluster b -- schema + migrations + kg cleanup **run date:** 2026-05-20\n**branch:** audit-w3-cluster-b (off origin/audit-weekly-2026-05-20)\nstatus convention: done = code landed; logged = called out for a follow-up.\n`db/migrations/0088_fk_supporting_indexes.sql`. audited every fk in docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-b-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-b-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-c-log.md", "kind": "doc", "title": "Wave-3 Cluster C execution log", "summary": "**Run date:** 2026-05-20\n**Branch:** `audit-w3-cluster-c` off `audit-weekly-2026-05-20`\n**Scope:** LLM tier diversification + dream-cycle concurrency + loop pull-through\n**Entries:** W3-5, W3-6, W3-7", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-c-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-c-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-3 cluster c execution log **run date:** 2026-05-20\n**branch:** `audit-w3-cluster-c` off `audit-weekly-2026-05-20`\n**scope:** llm tier diversification + dream-cycle concurrency + loop pull-through\n**entries:** w3-5, w3-6, w3-7 docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-c-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-c-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-d-log.md", "kind": "doc", "title": "Wave-3 Cluster D -- Frontend decomposition + graph API", "summary": "- **Branch:** `audit-w3-cluster-d` (off `audit-weekly-2026-05-20`)\n- **Worktree:** `/Users/rahulthayil/.superset/projects/torus/.claude/worktrees/agent-a0f7a2ff68d9d5c27`\n- **Run date:** 2026-05-20\n- **Commits:** 12", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-d-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-d-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-3 cluster d -- frontend decomposition + graph api - **branch:** `audit-w3-cluster-d` (off `audit-weekly-2026-05-20`)\n- **worktree:** `/users/rahulthayil/.superset/projects/torus/.claude/worktrees/agent-a0f7a2ff68d9d5c27`\n- **run date:** 2026-05-20\n- **commits:** 12 docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-d-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-3-cluster-d-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-a-log.md", "kind": "doc", "title": "Wave-4 Cluster A — Auth bypass paths", "summary": "Branch: `audit-w4-cluster-a` (off `audit-weekly-2026-05-20`).\nRun date: 2026-05-20.\nCluster scope: W4-1b, W4-1c, W4-1d, W4-1e, W4-1a (in that order).\n**Decision:** Option 1 + CI grep gate.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-a-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-a-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-4 cluster a — auth bypass paths branch: `audit-w4-cluster-a` (off `audit-weekly-2026-05-20`).\nrun date: 2026-05-20.\ncluster scope: w4-1b, w4-1c, w4-1d, w4-1e, w4-1a (in that order).\n**decision:** option 1 + ci grep gate. docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-a-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-a-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-c-log.md", "kind": "doc", "title": "Wave-4 Cluster C -- RLS + tool idempotency", "summary": "- **Branch:** `audit-w4-cluster-c` (off `audit-weekly-2026-05-20`)\n- **Worktree:** `/Users/rahulthayil/.superset/projects/torus/.claude/worktrees/agent-a0e5f5d5d0c54c607`\n- **Run date:** 2026-05-20\n- **Commits:** 3 (1x W4-7+W4-21, 1x W4-6+W4-23, 1x W4-43)", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-c-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-c-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-4 cluster c -- rls + tool idempotency - **branch:** `audit-w4-cluster-c` (off `audit-weekly-2026-05-20`)\n- **worktree:** `/users/rahulthayil/.superset/projects/torus/.claude/worktrees/agent-a0e5f5d5d0c54c607`\n- **run date:** 2026-05-20\n- **commits:** 3 (1x w4-7+w4-21, 1x w4-6+w4-23, 1x w4-43) docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-c-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-c-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-d-log.md", "kind": "doc", "title": "Wave-4 Cluster D — execution log", "summary": "**Branch:** `audit-w4-cluster-d` (base `audit-weekly-2026-05-20`)\n**Run date:** 2026-05-20\n**Entries:** W4-10, W4-11, W4-24, W4-35 (in that order)\n---", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-d-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-d-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-4 cluster d — execution log **branch:** `audit-w4-cluster-d` (base `audit-weekly-2026-05-20`)\n**run date:** 2026-05-20\n**entries:** w4-10, w4-11, w4-24, w4-35 (in that order)\n--- docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-d-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-d-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-g-log.md", "kind": "doc", "title": "Wave-4 Cluster G execution log", "summary": "**Run date:** 2026-05-20\n**Branch:** `audit-w4-cluster-g` off `audit-weekly-2026-05-20`\n**Scope:** LLM compaction threshold + per-call usage tracking — W4-27, W4-41\nTwo LLM-gateway changes. W4-27 is a one-line config bump (with the", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-g-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-g-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-4 cluster g execution log **run date:** 2026-05-20\n**branch:** `audit-w4-cluster-g` off `audit-weekly-2026-05-20`\n**scope:** llm compaction threshold + per-call usage tracking — w4-27, w4-41\ntwo llm-gateway changes. w4-27 is a one-line config bump (with the docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-g-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-g-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "doc:docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-h-log.md", "kind": "doc", "title": "Wave-4 Cluster H — Content security + sandbox", "summary": "**Branch:** `audit-w4-cluster-h`\n**Base:** `audit-weekly-2026-05-20`\n**Run date:** 2026-05-20\nFive entries: W4-32, W4-33, W4-34, W4-38, W4-40.", "path": "docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-h-log.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-h-log.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "audits", "2026-05-20-weekly" ], "searchText": "wave-4 cluster h — content security + sandbox **branch:** `audit-w4-cluster-h`\n**base:** `audit-weekly-2026-05-20`\n**run date:** 2026-05-20\nfive entries: w4-32, w4-33, w4-34, w4-38, w4-40. docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-h-log.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/audits/2026-05-20-weekly/wave-4-cluster-h-log.md github runbook active docs runbook docs superpowers audits 2026-05-20-weekly" }, { "id": "skill:weekly-risk-roundup", "kind": "autopilot-skill", "title": "Weekly Risk Roundup", "summary": "---\nname: weekly-risk-roundup\ndescription: |\nUse when producing a weekly blog post covering the most relevant global risk events.", "path": "ops/autopilot/skills/weekly-risk-roundup/SKILL.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/ops/autopilot/skills/weekly-risk-roundup/SKILL.md", "source": "AUTOPILOT", "status": "active", "tags": [ "autopilot", "skill", "skills", "weekly-risk-roundup" ], "searchText": "weekly risk roundup ---\nname: weekly-risk-roundup\ndescription: |\nuse when producing a weekly blog post covering the most relevant global risk events. ops/autopilot/skills/weekly-risk-roundup/skill.md https://github.com/torus-intelligence/torus/blob/staging/ops/autopilot/skills/weekly-risk-roundup/skill.md autopilot active autopilot skill skills weekly-risk-roundup" }, { "id": "doc:docs/operator-setup.md", "kind": "doc", "title": "What I need from you", "summary": "Cross-checked against Railway **staging** env. Already-set items are below for reference; the actionable list is at the top.\nAll credential capture is done; wiring it into a runtime is paused. Resume when a Codex seat is acquired, or when a specific autopilot skill becomes priority enough to justify the work.\n**When resumed:**\n- Codex env vars hold the secrets directly. Hardcoded per-environment paste-in. No external secret manager, no CLI integration.", "path": "docs/operator-setup.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/operator-setup.md", "source": "GITHUB", "category": "DEPLOYMENT", "status": "active", "tags": [ "docs", "deployment", "docs" ], "searchText": "what i need from you cross-checked against railway **staging** env. already-set items are below for reference; the actionable list is at the top.\nall credential capture is done; wiring it into a runtime is paused. resume when a codex seat is acquired, or when a specific autopilot skill becomes priority enough to justify the work.\n**when resumed:**\n- codex env vars hold the secrets directly. hardcoded per-environment paste-in. no external secret manager, no cli integration. docs/operator-setup.md https://github.com/torus-intelligence/torus/blob/staging/docs/operator-setup.md github deployment active docs deployment docs" }, { "id": "doc:docs/superpowers/plans/2026-05-22-workspace-resilience-and-dedup.md", "kind": "doc", "title": "Workspace Resilience + Content-Addressable Dedup Implementation Plan", "summary": "> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n**Goal:** Unblock prod workspace (PDFs stale on switch, NO TRACKED VALUES on Feather, \"NO DOCUMENT OPEN\" on read-only toggle) and build the foundation for content-addressable dedup so identical files across projects share extracted artifacts cleanly, without false attribution.\n**Architecture:**\n1. **Race fixes:** add cleanup guards to the three workspace `$effect` fetches that have no cancellation token — older promises stop clobbering newer state.", "path": "docs/superpowers/plans/2026-05-22-workspace-resilience-and-dedup.md", "githubUrl": "https://github.com/Torus-Intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-22-workspace-resilience-and-dedup.md", "source": "GITHUB", "category": "RUNBOOK", "status": "active", "tags": [ "docs", "runbook", "docs", "superpowers", "plans" ], "searchText": "workspace resilience + content-addressable dedup implementation plan > **for agentic workers:** required sub-skill: use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. steps use checkbox (`- [ ]`) syntax for tracking.\n**goal:** unblock prod workspace (pdfs stale on switch, no tracked values on feather, \"no document open\" on read-only toggle) and build the foundation for content-addressable dedup so identical files across projects share extracted artifacts cleanly, without false attribution.\n**architecture:**\n1. **race fixes:** add cleanup guards to the three workspace `$effect` fetches that have no cancellation token — older promises stop clobbering newer state. docs/superpowers/plans/2026-05-22-workspace-resilience-and-dedup.md https://github.com/torus-intelligence/torus/blob/staging/docs/superpowers/plans/2026-05-22-workspace-resilience-and-dedup.md github runbook active docs runbook docs superpowers plans" } ], "stats": { "total": 204, "docs": 145, "autopilotOutputs": 7, "runners": 12, "skills": 31 } }