done-when-criteria
Skill0 starsChanged 15 days agoMITView on GitHub
Define completion as observable artifacts so a skill cannot be declared finished on vibes. Use when a procedure has several steps and the endpoint of the work is ambiguous
In this repository49 other files it gives its agents.
- argument-construction-from-artifactsSkill
- audit-log-of-tool-callsSkill
- bounded-output-designSkill
- capability-negotiation-checkSkill
- changelog-and-versioningSkill
- concurrency-and-ordering-checkSkill
- context-window-managementSkill
- credential-isolationSkill
- data-retention-policySkill
- description-as-contractSkill
- destructive-action-confirmationSkill
- destructive-tool-gatingSkill
- deterministic-command-choicesSkill
- error-shape-for-agentsSkill
- fixture-driven-transcript-testsSkill
- human-in-the-loop-escalationSkill
- idempotency-annotationSkill
- input-schema-strictnessSkill
- least-privilege-tool-grantsSkill
- lifecycle-test-initialize-then-toolsSkill
- local-vs-remote-boundarySkill
- one-procedure-per-skillSkill
- pagination-and-limit-disciplineSkill
- parallel-tool-call-safetySkill
- protocol-conformance-testSkill
- protocol-version-handshakeSkill
- publishing-a-pack-locallySkill
- rate-limit-respectSkill
- reference-canonicalisationSkill
- replay-transcript-for-regressionSkill
- restart-and-reload-semanticsSkill
- scope-guard-before-executionSkill
- search-then-fetch-patternSkill
- secrets-via-environmentSkill
- server-discovery-and-trustSkill
- signal-and-pitfall-sectionsSkill
- skill-frontmatter-contractSkill
- stdio-client-configSkill
- stdout-purity-checkSkill
- third-party-server-sandboxingSkill
- tool-call-budgetingSkill
- tool-call-error-pathsSkill
- tool-error-without-retry-stormSkill
- tools-list-schema-validationSkill
- tool-surface-minimisationSkill
- tool-verb-namingSkill
- transport-choiceSkill
- verify-tool-output-before-reportingSkill
- when-to-use-and-avoidSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
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
Write what exists on disk when the skill is complete
- Distinctive: line 23
Avoid verbs like understand, consider or ensure
- Distinctive: line 24
Check each criterion can be marked pass or fail without argument
- Distinctive: line 30
no criterion depends on an unstated judgement
- Distinctive: line 62
every criterion is a pass or fail check on an artifact
≈ “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
Done When Criteria
Define completion as observable artifacts so a skill cannot be declared finished on vibes
Use when: a procedure has several steps and the endpoint of the work is ambiguous
Avoid when: the skill is one mechanical action whose result is self-evident
Procedure
- Write what exists on disk when the skill is complete
- Require at least one artifact a reviewer can open
- Include a verification step that compares two artifacts
- Avoid verbs like understand, consider or ensure
- Check each criterion can be marked pass or fail without argument
- Re-read after a session and confirm you actually produced each artifact
Signals
- each criterion names a file, a pair or a captured response
- no criterion depends on an unstated judgement
- the criteria can be checked by someone else
- a session that skipped a criterion has a visible hole
Patterns
1. Artifact criterion
a baseline and candidate capture exist for the same target2. Verification criterion
a second run reproduces the same digest3. Rejected verb
understand the cache behaviourPitfalls
- Writing completion criteria as feelings or intentions
- Requiring an artifact that the procedure never produces
- Making criteria so strict that routine work always fails them
Done when
- every criterion is a pass or fail check on an artifact
- a completed session can be audited against the list
References
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.

