Code Refactor
v0.1.0Improve code readability, structure, and maintainability by cleaning up duplication, simplifying flow, renaming, or reorganizing without changing behavior.
⭐ 0· 145·0 current·0 all-time
by@sf0799
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description (code refactor) match the SKILL.md instructions: identify duplication, simplify flow, rename, reorganize without changing behavior. No unrelated env vars, binaries, or installs are requested.
Instruction Scope
Instructions correctly focus on refactoring and include guidance to verify behavior remains unchanged. This necessarily implies reading and modifying the user's local codebase and running whatever tests or verification the user has — expected for a refactoring skill, but it means the agent may access project files and test harnesses. The SKILL.md does not prescribe how to run verification (e.g., run test suite, use sandbox), so the agent's actions for verification depend on agent implementation and must be authorized by the user.
Install Mechanism
No install spec or code files are present (instruction-only), so nothing will be written to disk by an installer. Lowest-risk install profile.
Credentials
No environment variables, credentials, or config paths are requested. The skill does not ask for unrelated secrets or system access.
Persistence & Privilege
Defaults are used (not always:true). The skill is user-invocable and can be invoked autonomously by the agent per platform defaults; this is normal and not excessive for this skill.
Assessment
This skill is internally consistent and appears to do what it says: read your project, propose and apply refactors, and verify behavior. Before using it: (1) ensure the agent has explicit permission to read and modify your repository; (2) keep a VCS commit or backup so you can review and revert changes; (3) confirm how the agent will verify correctness (e.g., will it run your test suite or rely on static checks), and restrict execution if you don't want tests run automatically; (4) require the agent to show diffs and ask for user confirmation before applying changes. If you are uncomfortable granting file-modification rights, do not enable autonomous invocation or run the skill only in a sandboxed environment.Like a lobster shell, security has layers — review code before you run it.
latestvk975vtsg9p431yefr4sqtmfp2584b2vh
License
MIT-0
Free to use, modify, and redistribute. No attribution required.