The Agent Skills Directory

[SAFE]: The skill uses local git operations and does not attempt to access sensitive files outside of the project scope or transmit data to external servers.
[SAFE]: Execution of critical actions (commit and push) is gated by a mandatory human confirmation step, which protects against unintended repository modifications.
[PROMPT_INJECTION]: The skill demonstrates an indirect prompt injection surface as it ingests untrusted data from local file diffs. While a malicious file could attempt to hijack the agent's context, the impact is constrained by the instructions and the final human review requirement.
Ingestion points: git diff output (analyzed uncommitted changes)
Boundary markers: None
Capability inventory: Shell access via Bash tool, file read/edit capabilities
Sanitization: None

git-commit