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 quarry-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/quarry-recall)<a href="https://agentmods.dev/skills/punt-labs/quarry/quarry-recall"><img src="https://agentmods.dev/badge/skills/punt-labs/quarry/quarry-recall/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/punt-labs/quarry/quarry-recall"><img src="https://agentmods.dev/badge/skills/punt-labs/quarry/quarry-recall.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00192 | $0.01033 |
| Opus 5.5 | $0.00077 | $0.00413 |
| Sonnet 5.5 | $0.00038 | $0.00207 |
| Haiku 4.5 | $0.00019 | $0.00103 |
Grade A, and why
quarry-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 24d 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 — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Quarry Recall
quarry find answers "has anyone already figured this out." Reach for it
before spending a WebSearch/WebFetch call, and before answering a
why/how/what-did-we-decide question from memory alone.
Recognize the trigger
- Understand code. "How does X work", "where is X defined", "what calls X", "explain this module."
- Recall project knowledge. "What did we decide about X", "why is it built this way", "did we already try Y."
- Recall memory. "What did I learn about X", "any gotchas here", "has a teammate hit this before."
All three are the same tool with different scope — pick the flags, not a different command.
Run it
quarry find "<natural-language question>"
Natural language beats keywords ("what did we decide about retry limits"
outperforms "retry limits"). Confirm the exact flag set with quarry find --help before relying on one not shown here — flags are the live contract,
this skill is not.
Scope the search to match the intent:
- Everything indexed:
quarry find "query"(no scope flags). - Only your own memories:
quarry find "query" --agent-handle <your-handle>. Always your own handle — never guess or borrow another agent's. - One collection:
quarry find "query" --collection <name>. - Narrow further:
--document,--page-type,--source-format,--memory-type(e.g.fact,procedure,lesson).
grep vs. find
| Need | Use |
|---|---|
| Exact symbol, literal string, a value you can name | grep — faster, exact, no embedding round-trip |
| Meaning, a paraphrase, "why"/"how"/"what did we decide" | quarry find |
| The file already open | The editor — don't search for what's on screen |
Read the result honestly
- Cite what you use:
[document p.N]. A claim without a citable hit is your own reasoning, not quarry's. - Empty is not "false." No results means quarry has not indexed it — say so plainly. Never present an empty search as proof something didn't happen or isn't true.
- Results are memory, dated. A hit reflects the state of the world when it was captured or indexed, not necessarily now. Check the date before treating a hit as current truth.
- You reasoned; quarry retrieved. Don't attribute a conclusion to quarry ("quarry says X") — quarry supplied the evidence, you drew the conclusion.
- No branded headings, emoji banners, or repeated summary blocks in the answer — plain prose citing sources.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 24d ago First seen · 102 lines · 192 tokens per session scan A 5a4998905182
quarry-recall is a skill published in the GitHub repository punt-labs/quarry (6 stars, last pushed yesterday), licensed MIT. It adds 192 tokens to every session and 1,033 once invoked, about $0.0008 per session on Opus 5.5. 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 2026-09-22.
Other skills, from other repositories
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.
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…
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…
stats
Show Captain Memo's corpus statistics (chunks per channel, observation counts, indexing progress, embedder info). Use when the user types /captain-memo:stats.
search
Hybrid search across Captain Memo's local memory + skills + observations. Use when the user types /captain-memo:search to retrieve top hits without the model having to decide whether to call searchall on its own.
doctor
Health probe across the worker, embedder, summarizer, queue, vector index, schema, cross-AI wiring and plugin install (cache and live versions). Use when the user types /captain-memo:doctor or when something seems wrong.