メニュー
Read-side helpers for Honcho memory and SpecSafe slice cost/activity. Status, review, cost, history, search — all read-only.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | memory |
| description | Read-side helpers for Honcho memory and SpecSafe slice cost/activity. Status, review, cost, history, search — all read-only. |
Provides read-only visibility into SpecSafe slice state and Honcho memory. No mutations are made to Honcho or to the state file.
HONCHO_API_KEY — required for review and search commandsHONCHO_WORKSPACE_ID — required for search; optional for review (falls back to workspaceId stored in state)Print a summary of the currently open slice, including cost counters.
bun run .omp/skills/memory/bin/memory.ts status
If no slice is open, prints "no slice currently open" and shows the last 3 history entries.
List all Honcho conclusions written during the given session, filtered to conclusions whose created_at falls within the session's beganAt..endedAt window (with a ±5 s grace band).
bun run .omp/skills/memory/bin/memory.ts review sess-xyz789
Iterates over the three conclusion-writer peers (validator, reviewer, steward) and aggregates results.
Print a human-readable cost breakdown for a slice.
bun run .omp/skills/memory/bin/memory.ts cost
bun run .omp/skills/memory/bin/memory.ts cost SPEC-20260424-003
If <slice-id> is omitted, uses the current open slice; if no slice is open, uses the most recent history entry.
Print recent finished slices with their outcome and cost. Default limit is 10.
bun run .omp/skills/memory/bin/memory.ts history
bun run .omp/skills/memory/bin/memory.ts history --limit=5
Semantic + text search across Honcho memory at workspace scope. Iterates over the conclusion-writer peers (validator, reviewer, steward) and merges results.
bun run .omp/skills/memory/bin/memory.ts search "login regression"
All output is plain text — no JSON, no color codes. Numbers are right-aligned in their columns. Tables include a total: footer where appropriate.
| Code | Meaning |
|---|---|
| 0 | Success |
| 1 | Usage error (wrong args, slice not found) |
| 2 | State or network error (missing env, Honcho) |
This skill carries inline copies of the statePathFor / readStateFileOrNull helpers from .pi/extensions/specsafe-session (memory also inlines CONCLUSION_WRITERS from .pi/extensions/honcho) so that it runs identically under pi and omp. Pin tests under .omp/test/specsafe.test.ts and .omp/test/honcho.test.ts enforce shape parity with the canonical extensions; see SPEC-008.2.
Initialize a foreign project to use the pi-seshat system. Symlinks .omp/, drops AGENTS.md and CLAUDE.md templates, scaffolds specs/ and .pi/, updates .gitignore. Dry-run by default; --i-approve gates all mutations.
Read-only pre-flight verifier. Checks Honcho/Linear/gh/omp connectivity, agent symlinks, and state-file integrity before a dogfood session. No mutations, no --i-approve gate.
Fetch, cache, and expose the latest official documentation for registered libraries, SDKs, and APIs. Turns Luci's "always check latest official docs" rule from prose into an executable contract.
Fetch, cache, and expose the latest official documentation for registered libraries, SDKs, and APIs. Turns Luci's "always check latest official docs" rule from prose into an executable contract.
Propose-review-apply flow for BMad artifacts (PRDs, UX specs, architecture, briefs). All applies require --i-approve and produce commits with Proposed-By/Approved-By trailers.
Read and draft-mutate GitHub PRs, issues, and API endpoints via gh(1). All mutations require --i-approve. PR creation enforces a Linear-state invariant and is logged to .pi/.github-log.jsonl.