The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the dynamic context injection syntax (!command) to gather repository metadata (such as remote URLs and branch status) at load time. These commands are benign git operations used solely for providing project context to the agent.
[COMMAND_EXECUTION]: The skill utilizes authorized development tools including gh, npm, and uv to interact with GitHub and execute local code quality scripts. These tools are used according to their intended purpose within the developer workflow.
[PROMPT_INJECTION]: The skill identifies a potential surface for indirect prompt injection through the ingestion of external data. 1. Ingestion points: The skill reads external data from GitHub PR logs via gh pr checks and repository files via the Read tool. 2. Boundary markers: No explicit delimiters or instructions are used to separate untrusted log content from the agent's instructions. 3. Capability inventory: The skill can execute shell commands through Bash and modify project files using the Edit tool. 4. Sanitization: No explicit validation or filtering of external log data is performed before analysis. This finding is documented as a structural risk inherent to PR analysis tasks.

git-fix-pr