Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/shankarr1503-safe-refactor
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/safe-refactor
What's inside
- Skill name
safe-refactor- Description
- Restructure code without changing behavior. Use when asked to refactor, clean up, simplify, rename, extract, split a large file or function, or remove duplication. Locks behavior with tests first, then moves in small verified steps.
- License
- MIT
- Instructions
- 39 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Safe refactor A refactor changes structure, never behavior. If behavior changes, it is not a refactor — it is a rewrite with extra risk. ## 1. Lock behavior first - Find the tests covering the code. Run them; they must be green before you start. - If coverage is thin, add characterization tests that pin the current behavior — including odd behavior that callers may depend on. (See the `legacy-tests` skill.) - Find every caller: search for the symbol, dynamic uses (string names, reflection, routes, DI config, templates), exports, and public API consumers outside the repo. ## 2. Move in small steps Each step is one mechanical transformation, followed by running the tests: - Rename (use the IDE/language server rename where possible, then search for string references). - Extract function / module; inline function; move file (update imports). - Replace conditional with lookup/polymorphism; introduce parameter object. - Delete dead code — only after proving it is unused (search, coverage, logs). "Looks unused" is not proof for public or dynamically-referenced code. Commit (or checkpoint) after each green step so you can bisect or revert. ## 3. Rules - **No behavior changes mixed in.** Found a bug? Note it, finish or pause the refactor, and fix the bug in a separate change with its own test. - **No formatting churn** in files you are not otherwise changing. - **Keep public interfaces stable** unless changing them is the goal; if so, migrate callers in the same change or leave a deprecated shim. - **Don't over-abstract.** Three similar blocks are fine; an abstraction must make the code easier to read _now_, not hypothetically flexible. - Performance-sensitive code: benchmark before and after. ## 4. Verify - Full test suite, typecheck, lint pass. - The diff reads as pure structure: a reviewer can confirm no behavior changed. Large moves are easier to review as separate "move only" and "edit" commits. ## Report What was restructured and why it is better (concretely: fewer branches, smaller function, removed duplication of X), the tests that guard it, and any bugs or follow-ups noticed but deliberately not fixed.
From shankarr1503/omnistack, MIT.
More from shankarr1503/omnistack
- 0.565score
api-contract
shankarr1503
Design or review an API (REST, GraphQL, gRPC, SDK or library interface) for clarity, consistency and long-term compatibility.
SkillAFreshness: AClaude Code +9 - 0.565score
commit-craft
shankarr1503
Write clean commits and pull request descriptions that reviewers love.
SkillAFreshness: AClaude Code +9 - 0.565score
fix-ci
shankarr1503
Diagnose and fix a failing CI pipeline, build or test job.
SkillAFreshness: AClaude Code +9 - 0.565score
legacy-tests
shankarr1503
Add tests to untested or legacy code before changing it.
SkillAFreshness: AClaude Code +9 - 0.565score
map-codebase
shankarr1503
Rapidly understand an unfamiliar repository and produce a practical map of it.
SkillAFreshness: AClaude Code +9 - 0.565score
perf-hunt
shankarr1503
Measure-first performance optimization.
SkillAFreshness: AClaude Code +9 - 0.565score
root-cause
shankarr1503
Find the real cause of a bug before changing code.
SkillAFreshness: AClaude Code +9 - 0.565score
safe-migration
shankarr1503
Plan and write zero-downtime database and schema migrations.
SkillAFreshness: AClaude Code +9 - 0.565score
ship-review
shankarr1503
Rigorous pre-merge review of a diff, branch or pull request.
SkillAFreshness: AClaude Code +9 - 0.565score
skill-forge
shankarr1503
Create a new high-quality agent skill (SKILL.md) for Claude Code, Codex or OpenCode.
SkillAFreshness: AClaude Code +9 - 0.565score
threat-check
shankarr1503
Security review of code or a change for exploitable vulnerabilities.
SkillAFreshness: AClaude Code +9 - 0.565score
upgrade-deps
shankarr1503
Upgrade dependencies, frameworks or language versions safely.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 7, 2026 (22 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/shankarr1503/omnistack
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from shankarr1503
- 0.565score
commit-craft
shankarr1503
Write clean commits and pull request descriptions that reviewers love.
SkillAFreshness: AClaude Code +9 - 0.565score
fix-ci
shankarr1503
Diagnose and fix a failing CI pipeline, build or test job.
SkillAFreshness: AClaude Code +9 - 0.565score
legacy-tests
shankarr1503
Add tests to untested or legacy code before changing it.
SkillAFreshness: AClaude Code +9 - 0.565score
map-codebase
shankarr1503
Rapidly understand an unfamiliar repository and produce a practical map of it.
SkillAFreshness: AClaude Code +9 - 0.565score
perf-hunt
shankarr1503
Measure-first performance optimization.
SkillAFreshness: AClaude Code +9 - 0.565score
api-contract
shankarr1503
Design or review an API (REST, GraphQL, gRPC, SDK or library interface) for clarity, consistency and long-term compatibility.
SkillAFreshness: AClaude Code +9
Cite this page
safe-refactor skill. Agent Mart. https://agentmart.dev/skills/shankarr1503-safe-refactor (last updated 2026-10-24)