The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git interpret-trailers, git log, and git config tools alongside standard utilities like grep, sort, uniq, and cut as defined in the allowed-tools and command examples in SKILL.md.
[PROMPT_INJECTION]: (1) Ingestion points: Untrusted data enters the agent context via the output of git log commands. (2) Boundary markers: There are no specific delimiters or instructions provided to the agent to disregard embedded commands within commit messages. (3) Capability inventory: The skill allows for reading repository history and modifying configuration settings. (4) Sanitization: No content-based filtering or sanitization is applied to the metadata extracted from commit logs.

git-commit-trailers