The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from git commit logs and diffs, which could contain malicious instructions from external contributors intended to influence the agent's output.
Ingestion points: git log and git diff output processed in Steps 2 and 3.
Boundary markers: Absent; the skill does not use specific delimiters to separate commit data from instructions.
Capability inventory: Read-only git commands and text generation; no file system write or network capabilities are present in the skill scripts.
Sanitization: Absent; commit messages are processed as-is to generate the summary.
[COMMAND_EXECUTION]: The skill executes local shell commands to gather branch information and repository status.
Evidence: Uses git remote, git log, git branch, and git diff commands to collect data for analysis.

suggest-pr