The Agent Skills Directory

[SAFE]: The skill's primary function is code cleanup. It provides a structured confidence scoring system (0-100) and specific examples of patterns to remove (extra comments, defensive checks, silent try/catch), which reduces the risk of unintended modifications.
[SAFE]: The operations performed by the skill, such as git diff and file reading/writing, are standard for developer productivity tools and are executed locally within the user's project context.
[SAFE]: The skill includes a clear 'False Positives' list and instructs the agent to run a linter or formatter after changes, following industry best practices for automated code editing.
[SAFE]: While the skill processes untrusted input (code diffs), the instructions constrain the agent to pattern-based removal, effectively mitigating the risk of indirect prompt injection from malicious code comments.

deslop