The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes untrusted data from staged git changes to generate commit messages.
Ingestion points: The workflow requires reading staged code changes to determine the nature of the modification as described in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore instructions embedded within the code changes are included in the instructions or templates.
Capability inventory: The skill is limited to text generation for commit messages. There are no indications of subprocess execution, file system modifications, or network operations within the provided files.
Sanitization: The skill does not perform sanitization or validation on the input code changes before incorporating them into the reasoning process for message generation.

conventional-commits