close-change
Skill3 starsChanged 8 days agoMITView on GitHub
Judge whether a verified change is ready for handoff, merge, release, or cleanup, or whether an explicitly guarded recovery action may proceed. Not for reporting an implementation run's own outcome, which implement-change owns, or for read-only branch facts, which analyze-project answers.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- docs/AGENTS.mdAGENTS.md
- analyze-projectSkill
- api-contract-strategySkill
- architecture-patternsSkill
- design-changeSkill
- development-standardsSkill
- docker-multiarch-buildSkill
- error-patternsSkill
- go-guidelinesSkill
- herdr-implementSkill
- implement-changeSkill
- logging-standardsSkill
- lua-guidelinesSkill
- organize-docsSkill
- output-stylesSkill
- plan-changeSkill
- powershell-guidelinesSkill
- python-guidelinesSkill
- review-changeSkill
- review-designSkill
- review-implementationSkill
- review-planSkill
- routeros-scripting-guidelinesSkill
- runtime-diagnosisSkill
- security-guardrailsSkill
- shell-guidelinesSkill
- skill-authoringSkill
- skill-minerSkill
- skills-routingSkill
- smart-commitSkill
- sops-age-guardrailsSkill
- sync-truthSkill
- testing-strategySkill
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
Check implementation outcome, verification, the implementation review when one was required, accepted-repair evidence, truth-sync status, …
≈ “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 Change
Decide whether a bounded change is complete for the requested close purpose.
Use This Skill When
- implementation and required verification are complete
- the user wants an explicit handoff, merge, release, cleanup, or closure judgment
Discovering a missing review, truth sync, or authority at closure produces the matching needs- outcome. A standalone request to remove, prune, or repair a Git worktree uses references/worktree-isolation.md in implement-change.
Closure Judgment
- Confirm the requested close purpose and the evidence relevant to it.
- Check implementation outcome, verification, the implementation review when one was required, accepted-repair evidence, truth-sync status, and any explicitly requested external action.
- Separate completed work from actions not authorized or not performed (commit, push, publication, deployment, merge, destructive cleanup). Name the delivered version, any authorized secondary tradeoff with its reason and impact, and what was not verified.
- Treat task membership, covered authority, and current capability as independent facts: an action outside this task stays out even under standing permission; an in-task action without covered authority returns
needs-decision; authority without capability or evidence returnsblocked. Consume already covered permission for the same target and side effects without re-asking, and pause only the increment whose goal, side effects, protected state, or permission changed. - Return
closedonly when all evidence required for the purpose is current for the specified version and endpoint; otherwise name the smallest owning gap. Disclosed authorized secondary tradeoffs do not blockclosedwhen the main goal and approved acceptance hold; an affected production action that was not performed leaves that endpoint incomplete.
Judge delivery by current evidence for the specified version at the requested endpoint rather than local checks or a delegated summary. A push carries every still-unpushed commit and the CI/CD it triggers; name that gap when it is outside covered authority.
Account for active delegated work, unintegrated results, pending review or repair, and task workspaces still needed; mark owned execution resources as reclaimed, retained for a stated need, or unresolved. The implementation owner performs cleanup of resources it owns; closure gains no mutation authority over user branches or other workspaces.
Outcomes
closed: the requested change boundary is complete for the stated purposeneeds-implementation: implementation or verification remains incompleteneeds-review: a required bounded review is absent or unresolvedneeds-truth-sync: stable truth still needs an authorized updateneeds-decision: the requested close action needs user or external authorityblocked: required evidence or capability is unavailable
Closure is a judgment, not permission to merge, release, delete, commit, push, publish, or deploy.
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.

