The Agent Skills Directory

[SAFE]: The skill follows security best practices by implementing a 'least privilege' approach to code modification. Instructions explicitly prohibit high-risk operations such as installing new dependencies, automatically committing/pushing code changes, or exfiltrating data. It includes specific guidance on verifying external code suggestions and ensuring that fixes are minimal and targeted to the authorized scope. The workflow requires manual verification steps and re-review of changes, which prevents the accidental introduction of malicious patterns from external suggestions.

codereview-fix-verify