The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code content from the repository and has the capability to execute commands and commit changes.
Ingestion points: The skill reads code changes using git diff, git diff --cached, and git status in SKILL.md (Step 1).
Boundary markers: Absent. The instructions do not define delimiters or specific constraints to ignore instructions that might be embedded within code comments or string literals.
Capability inventory: The skill has access to shell execution for running git commands, project-specific formatting/linting tools, and git commit (Steps 1, 2, 7, and 8).
Sanitization: Absent. The agent reviews raw code differences without sanitization or input validation of the content.
[COMMAND_EXECUTION]: The skill executes shell commands to analyze the repository state and perform code quality tasks.
Evidence: SKILL.md instructs the agent to run git diff, git status, git commit, and various formatting or linting commands tailored to the project's environment.

review-code