The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data that could contain malicious instructions. Ingestion points: Reads code changes via git diff and git status commands specified in SKILL.md. Boundary markers: No explicit delimiters are used to separate the diff content from the instructions. Capability inventory: The skill executes local git commands but does not have network access or file-write capabilities. Sanitization: None present. The risk is mitigated by the instruction to return only the final commit message with no extra text.
[COMMAND_EXECUTION]: The skill utilizes local git commands (git status, git diff) to gather context from the repository. These commands are used for inspection purposes and do not involve writing to the filesystem or executing external scripts.

write-conventional-commit