The Agent Skills Directory

[COMMAND_EXECUTION]: Employs git status and git diff --name-only to analyze repository changes. These are standard, safe commands for a development-focused skill.
[CREDENTIALS_UNSAFE]: Encourages the use of placeholders for sensitive data in documentation, reducing the risk of accidental credential exposure in public changelogs.
[INDIRECT_PROMPT_INJECTION]: While the skill processes repository metadata (file names), the risk of malicious instructions via file names is mitigated by the requirement for human confirmation before any generated text is committed.

gen-changesets