deep-code-cleanup
Skill1 starsChanged 23 days agoApache-2.0View on GitHub
Perform an exhaustive, iterative cleanup of a requested scope while preserving behavior.
In this repository25 other files it gives its agents.
- AGENTS.mdAGENTS.md
- add-testSkill
- audit-fixissuesSkill
- autoresearchSkill
- benchmark-optimizeSkill
- circleci-job-resultsSkill
- citation-verifierSkill
- cli-designSkill
- codebase-auditSkill
- code-cleanupSkill
- doc-updateSkill
- emendSkill
- fix-issueSkill
- git-github-workflowSkill
- goal-modeSkill
- inspect-datasetSkill
- manage-pr-lifecycleSkill
- medcognetics-brandSkill
- notify-wakeSkill
- repo-maintenanceSkill
- review-fix-loopSkill
- standardize-ciSkill
- token-file-authSkill
- whats-nextSkill
- yeetSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
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 18
Use the merge base for the comparison.
- Distinctive: line 31
Ask only when a necessary change exceeds the accepted scope or requires a material behavior decision.
- Distinctive: line 34
Add characterization coverage before refactoring weakly tested behavior.
- Distinctive: line 44
Run required repository gates on the final state.
≈ “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
Deep Code Cleanup
Accept $deep-code-cleanup, optional paths or functions, and aggressive for
repository-wide scope even when active changes exist.
Select the Scope
Explicit targets include their affected tests and documentation. Otherwise:
- On a feature branch, inspect the branch diff plus relevant working changes.
- On
mainormaster, use active changes, or the repository when clean. - Resolve the base from
origin/main,origin/master,main, thenmaster. Use the merge base for the comparison.
Exclude unrelated untracked packages, generated files, caches, and vendored trees unless requested. Preserve unrelated work.
Complete the Cleanup
Inspect the whole scope for concrete maintenance costs. Group shared causes, such as duplicated policy, tangled responsibilities, or brittle tests. Describe substantial changes in a progress update and proceed with authorized work.
Preserve behavior, public interfaces, dependency versions, and data formats. Report design problems with evidence and continue safe local improvements. Ask only when a necessary change exceeds the accepted scope or requires a material behavior decision.
Add characterization coverage before refactoring weakly tested behavior. For an authorized bug fix, confirm a failing regression first. Make coherent passes and run the smallest useful validation after each. Reinspect the diff for accidental behavior changes and new complexity.
Continue until no concrete, safe, in-scope improvements remain. Do not stop at the first refactor or iterate on subjective style changes. Resolve routine tooling and test failures caused by the work. Report external blockers and continue independent work.
Run required repository gates on the final state. Report the comparison base, improvements, behavior-preservation evidence, validation, and deferred decisions.
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.

