The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes standard git commands (git log, git diff, git show) to retrieve information about recent code changes from the local repository.
[SAFE]: No network operations, credential access, or persistent system modifications were found in the skill's instructions or reference files.
[SAFE]: The skill presents an indirect prompt injection surface by reading user-controlled commit messages and code diffs. However, this is considered safe because the skill lacks capabilities to perform sensitive actions such as network requests or file writes. Ingestion points: git log and diff outputs; Boundary markers: None; Capability inventory: Read-only git operations and limited file reads; Sanitization: None.

lesson-learned