Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add punt-labs/quarry --skill recallgit clone --depth 1 https://github.com/punt-labs/quarryWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/punt-labs/quarry/recall)<a href="https://agentmods.dev/skills/punt-labs/quarry/recall"><img src="https://agentmods.dev/badge/skills/punt-labs/quarry/recall.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00155 | $0.01064 |
| Opus 5 | $0.00077 | $0.00532 |
| Sonnet 5 | $0.00031 | $0.00213 |
| Haiku 4.5 | $0.00015 | $0.00106 |
Grade A, and why
recall scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Quarry — Local Semantic Search
Quarry indexes documents by meaning and answers natural-language questions against them: this repo's source and docs, prior session transcripts, and web pages fetched during earlier research. Reach for it before spending a WebSearch or WebFetch call re-discovering something already found.
When to use it
- Use find before WebSearch or WebFetch for research, or before answering a why/how/what-did-we-decide question. Quarry indexes this codebase, design docs, prior session transcripts, and previously fetched web pages — it often already has the answer.
- Prefer grep for symbol and value lookups; prefer find for meaning.
- Pick the capture verb by the shape of what you're saving. The four verbs
are distinct on purpose:
remember— a specific durable fact (a URL, an ID, an address, a version pin). Small, factual, retrievable by its literal content.ingest— a URL. Fetches the page with smart sitemap discovery and single-page fallback. For local files or directories, useregister_directory+sync_all_registrationsinstead.learn— a distilled lesson: the rule you'd tell a teammate ("when X, do Y, because Z"). Lessons route to the repo's-lessonscollection and get retrieval preference over transcripts and general docs, so a lesson typically ranks above a session transcript that mentions the same keywords.find— the retrieval verb; searches everything the other three wrote.
When not to use it
- Exact symbol or value lookups (a function name, a literal string) — grep is faster and precise.
- Navigating the file currently open — use the editor, not search.
- For architecture decisions inside this repo, read DESIGN.md directly — quarry find complements it, does not replace it. DESIGN.md is the authoritative ADR log; find shines when the question spans prior session transcripts, web-fetched research, or docs outside the current tree.
Tools
/find <query>— search the knowledge base; natural language beats keywords ("What did we decide about retry limits?" beats "retry limits")./remember <name>— persist inline text as a named memory (a durable fact)./learn <lesson>— save a distilled lesson that gets retrieval preference over general docs and transcripts./ingest <url>— fetch and index a URL (sitemap discovery with single-page fallback). For local files or directories, useregister_directory+sync_all_registrationsinstead./explain <document or topic>— search and synthesize an explanation./source <claim or text>— find which document a claim came from.- MCP tools (same operations, callable directly):
find,remember,learn,ingest,register_directory,sync_all_registrations,show,delete,list,status,use. Prefer/quarry:quarry use <db>as the interactive entry for database switching; theuseMCP tool is also available.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago Changed · +25 lines · +6 tokens per session 991623766440
- 7d ago First seen · 61 lines · 149 tokens per session scan A 03420a71ad4e
recall is a skill published in the GitHub repository punt-labs/quarry (3 stars, last pushed 3d ago), licensed MIT. It adds 155 tokens to every session and 1,064 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
memem-mine
Opt in to memem's event-triggered mining. New sessions are mined automatically via the Stop hook.
pond
Recall and analyze past AI agent sessions (Claude Code, Codex, opencode, and more). Find prior work and decisions, read/review/summarize a past session transcript, or run SQL analytics over session history. Use whenever the user references past sessions, prior work, "check pond", or asks what was done or decided…
autosearch:context-retention-policy
Session-level policy for keeping the runtime AI's context window healthy across long research — keep-last-k tool results, offload older evidence to disk, trigger compaction at thresholds. Borrows MiroThinker's keeptoolresult, deepagents' summarization middleware, and deer-flow's SummarizationEvent pattern. Orthogonal…
autosearch:experience-compact
Promote recurring patterns from experience/patterns.jsonl into the compact experience.md digest (≤120 lines, read by runtime AI before calling the skill). Triggers on N-events / file-size / user-feedback / session-end. Guards against single-success noise and pollution via promotion thresholds.
stats
Show Captain Memo's corpus statistics (chunks per channel, observation counts, indexing progress, embedder info). Use when the user types /captain-memo:stats.
observations
List recent captured session observations (the Haiku-summarized voyage logs). Use when the user wants to see what Captain Memo has logged from past sessions.