- name
SKILL.md
Le fichier du skill reste dans le dépôt de l'éditeur.
Collez la commande copiée dans un terminal. Elle installe ce SKILL.md dans cet agent.
npx skills add omagents/omagents --skill refactor -a cursor -yCollez la commande copiée dans un terminal. Elle installe ce SKILL.md dans cet agent.
npx skills add omagents/omagents --skill refactor -a claude-code -yCollez la commande copiée dans un terminal. Elle installe ce SKILL.md dans cet agent.
npx skills add omagents/omagents --skill refactor -a codex -yCollez la commande copiée dans un terminal. Elle installe ce SKILL.md dans cet agent.
npx skills add omagents/omagents --skill refactor -a grok -yEnvoyez le message copié dans un chat Grok Bot. Le Bot récupère ce SKILL.md sur GitHub et l'ajoute aux skills que partagent tous vos Bots.
Install this skill: https://github.com/omagents/omagents/tree/HEAD/packages/core/skills/refactorDe l'éditeur
Systematically refactor code targets one by one, with verification after each change. The loop engine tracks progress and survives context clearing.
Scan the codebase for refactoring opportunities:
# Long functions (> 50 lines)
grep -rn "def \|function \|const .* = " --include="*.py" --include="*.ts" --include="*.js" | head -50
# High complexity (deep nesting)
# Use codegraph to find complex call paths
codegraph_codegraph_explore query="complex functions with deep nesting"
# Code duplication
# Use grep to find repeated patterns
grep -rn "duplicate pattern" --include="*.py"
loop_engine.py init refactor '[
{"file": "src/auth.py", "target": "split authenticate() into validate_credentials() + create_session()", "description": "Refactor auth.py: split authenticate()"},
{"file": "src/cache.py", "target": "replace manual caching with functools.lru_cache", "description": "Refactor cache.py: use lru_cache"},
{"file": "src/utils.ts", "target": "extract duplicate validation logic into shared validator", "description": "Refactor utils.ts: extract validator"}
]'
Repeat until next returns null:
Step 1: Get next task
loop_engine.py next refactor
If output is null, go to Phase 4.
Step 2: Read and understand the target
Read the file, understand what it does currently, and plan the specific changes.
Step 3: Apply the refactoring
Use the host's edit tools to make the changes. If the active host exposes LSP rename support:
lsp tool for prepare_rename and rename operationslsp tool for findReferences to ensure all callers are updatedIf using ast-grep:
sg -p 'old_pattern' -r 'new_pattern' --lang python --update-all <file>
Step 4: Verify
# Run tests
npm test 2>/dev/null || pytest tests/ 2>/dev/null || true
# Run linter
ruff check <file> 2>/dev/null || npx eslint <file> 2>/dev/null || true
# Check LSP diagnostics if available
# Use lsp tool with diagnostics operation
Step 5: Record result
If verification passes:
loop_engine.py complete refactor <id> "Split authenticate() into 2 functions, all tests pass"
If tests fail:
loop_engine.py fail refactor <id> "test_auth.py failed after split"
loop_engine.py summary refactor
Present the summary. For blocked tasks, show what failed and suggest manual review.
Systematic code refactoring using a durable loop workflow. Each refactoring target is processed, verified, and marked complete before moving to the next. Trigger when the user says 'refactor', 'restru
---name: refactordescription: Systematic code refactoring using a durable loop workflow. Each refactoring target is processed, verified, and…---# packages/core/skills/refactor/SKILL.mdVue d'ensemble
Systematic code refactoring using a durable loop workflow. Each refactoring target is processed, verified, and marked complete before moving to the next. Trigger when the user says 'refactor', 'restructure code', or wants to improve code without changing behavior.
FAQ
Téléchargez packages/core/skills/refactor/SKILL.md auprès de l'éditeur, ou copiez le chemin et ouvrez le dépôt. PlugMCP répertorie le skill et n'héberge pas le fichier.
L'éditeur le maintient sur https://github.com/omagents/omagents.
Claude, Cursor, Codex et Grok Build peuvent lire un SKILL.md que vous ajoutez depuis le dépôt de l'éditeur. Grok Bot l'installe lorsque vous envoyez dans un chat le lien du dossier GitHub du skill.