attic-recall
Skill5 starsChanged 2 months agoMITView on GitHub
Pull ONE item back out of .attic/ by slug or search words and summarise it. Use when the question names a TOPIC: "what did we find about X", "recall X", "what's in the attic about X", or when the user references an attic:<slug> handle. A question with no topic that just asks what the attic contains goes to attic-index instead.
In this repository12 other files it gives its agents.
- attic-doctorSkill
- attic-failSkill
- attic-gitSkill
- attic-helpSkill
- attic-indexSkill
- attic-pinSkill
- attic-pruneSkill
- attic-reviewSkill
- atticSkill
- attic-stashSkill
- attic-statsSkill
- attic-sweepSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 33
Never rewrite or delete items while recalling.
- Very specific: line 54
If the user acts on a stale item, offer to verify it against the current code and record the result with /attic-review.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Recall from the attic
Query: $ARGUMENTS.
node "${CLAUDE_SKILL_DIR}/../attic/scripts/attic.js" recall "$ARGUMENTS"The script scores slug, index hook and item body, returns the best match, and lists alternatives.
- No
.attic/yet, or nothing matches: say so in one line and stop. Do not invent an answer or go read source files unless the user asks. - Several plausible matches: report the best one, then list the alternative handles in one line so the user can redirect.
- Report the item in at most ten lines. Keep code, paths and commands verbatim. End with the handle.
Never rewrite or delete items while recalling.
Freshness
Recall prints a trust line above the body, and for anything that is not current, a warning with the reason. Pass it on — an item's staleness is part of the item, and dropping it while quoting the conclusion is the failure this release exists to prevent.
| What recall says | What you do |
current | Use it. Mention the item is current only if asked. |
possibly-stale | Quote it, then say which files changed and that it needs checking before it is relied on. |
needs-review | As above, and do not present the claim as fact. A cited file may be gone. |
unverified | Say it was recorded but never checked. |
| no trust line at all | An item from before 1.6. Treat its provenance as unknown, not as verified. |
The distinction to hold on to: possibly-stale means a file underneath this
finding moved, not that the finding is false. Report it as something to
check, never as something that is now wrong — you have not checked it either.
If the user acts on a stale item, offer to verify it against the current code
and record the result with /attic-review. Recall itself never writes.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

