The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local git commands including status, diff, log, and commit within a shell environment to manage the repository state.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests data from git diffs and logs which could contain malicious instructions. Evidence: 1. Ingestion points: The skill reads output from git diff, git diff --staged, and git log --oneline -5 in SKILL.md. 2. Boundary markers: No explicit markers are used to isolate the diff content from instructions. 3. Capability inventory: The skill has the ability to execute git commands on the local system. 4. Sanitization: A mandatory user review step is provided to confirm the generated commit message before it is finalized.

semantic-commit