# Refactor

> >

- **Type:** Skill
- **Install:** `agentstack add skill-ao92265-claude-code-playbook-refactor`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [ao92265](https://agentstack.voostack.com/s/ao92265)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [ao92265](https://github.com/ao92265)
- **Source:** https://github.com/ao92265/claude-code-playbook/tree/main/skills/refactor

## Install

```sh
agentstack add skill-ao92265-claude-code-playbook-refactor
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Focused Refactor

Improve code structure without changing behavior. Zero behavior change is non-negotiable.

## Steps

1. **Establish green baseline:**
   - Run the full test suite and confirm all tests pass
   - Run the type checker (`tsc --noEmit`, `go vet`, `cargo check`, `mypy`)
   - If tests don't exist for the code being refactored, stop and suggest using `/test-first`
   - Record the test results as the baseline

2. **Identify the refactoring:**
   - Clarify what the user wants to improve (or suggest based on code smells)
   - Common refactorings: rename, extract function/method, inline, simplify conditionals,
     remove duplication, split file, consolidate imports
   - Scope the change — one refactoring at a time

3. **Apply the refactoring:**
   - Make the minimum changes necessary
   - Update all references (imports, call sites, type definitions)
   - Preserve all public API contracts
   - Don't change test assertions — tests are the behavior contract

4. **Verify behavior is preserved:**
   - Run the full test suite again
   - Run the type checker again
   - Compare results to baseline — must be identical pass/fail

5. **If tests fail — revert immediately:**
   - `git checkout -- .` to discard changes
   - Analyze why the refactoring broke tests
   - Try a different, safer approach
   - Never fix tests to accommodate refactoring — that means behavior changed

6. **Final checks:**
   - Run linter to confirm no new warnings
   - Show the diff summary to the user
   - Confirm the refactoring achieved the intended improvement

## Important

- **Zero behavior change is the rule.** If any test fails after refactoring, revert. No exceptions.
- **Don't combine refactoring with feature work.** Refactor in one commit, features in another.
- **Tests are the contract.** If you need to change tests, you're not refactoring — you're changing behavior.
- **One refactoring per pass.** Don't rename AND extract AND simplify in one go.
- **If no tests exist, stop.** Refactoring without tests is guessing. Write tests first.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [ao92265](https://github.com/ao92265)
- **Source:** [ao92265/claude-code-playbook](https://github.com/ao92265/claude-code-playbook)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-ao92265-claude-code-playbook-refactor
- Seller: https://agentstack.voostack.com/s/ao92265
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.