The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands (git diff, git log, git rev-parse) to analyze the repository state. It validates user-provided references using git rev-parse before generating diffs to ensure inputs are valid repository objects.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from repository diffs and commit messages. * Ingestion points: git diff and git log output in SKILL.md. * Boundary markers: Sub-agent prompts in SKILL.md define narrow tasks to prevent context pollution. * Capability inventory: Read-only git operations for analysis. * Sanitization: None. The risk is mitigated by the constrained, parallel nature of the sub-agents which restricts the influence of instructions potentially embedded in code diffs.- [SAFE]: The skill implements a legitimate developer workflow using isolated sub-agents and standard repository analysis tools provided by the vendor.

code-review