close-the-books
Skill1 starsChanged 5 months agoMITView on GitHub
Before calling work done, reconcile it against its opening commitments — every acceptance criterion ticked, every invariant re-verified — with evidence, not assertion. Use when about to claim a task is complete or before merging.
In this repository8 other files it gives its agents.
- expert-panelSkill
- gather-contextSkill
- scan-projectSkill
- scope-watchSkill
- seal-planSkill
- servo-flowSkill
- servo-gateSkill
- shape-specSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are 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.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Account for every item: Every criterion ticked, every invariant verified, each with its evidence.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
close-the-books
Do not declare work done; reconcile it. Like counting the instruments before closing — the count at the end must match the commitments at the start, or something was left inside. This is a conservation check: the artifact's opening inventory (its acceptance criteria and the invariants it promised not to break) must all be accounted for, with evidence.
The failure this prevents: an agent asserting "done / fixed / passing" with confidence it has not earned, because it never ran the check.
Procedure
- Recover the opening inventory. From the plan and the manifest: the
acceptancecriteria (what "done" was defined as) and theinvariantsthe work promised not to break. - Tick each acceptance criterion — with evidence. For every criterion, run the actual check (the test, the command, the observation) and confirm it passes. Quote the evidence. A criterion with no run is not ticked.
- Re-verify each invariant. Confirm nothing the work promised not to break is broken. Run the relevant gate (
servo-gateagainstinvariants) if judgment is needed; otherwise check directly. - Account for every item. Every criterion ticked, every invariant verified, each with its evidence. If any item cannot be accounted for, the books do not close — the work is not done. Report the open item plainly.
Rule
Evidence before assertion, always. "It should work" is not closing the books. Run the check, quote the output, then close. If a check failed or was skipped, say so — do not round up to "done".
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.