The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes standard development tools and CLI commands to validate code and manage the PR lifecycle.
Validation tools include: pnpm, npm, mypy, pytest, ruff, and make for running builds, type checks, and tests.
GitHub management is handled via the official GitHub CLI (gh), including status checks, viewing comments, and merging PRs.
These commands are restricted to the local repository context and the authenticated user's GitHub permissions.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection, as it is designed to ingest data from external sources.
Ingestion points: The skill reads PR comments using the gh pr view <number> --comments command as described in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the contents of PR comments as untrusted data or to ignore embedded instructions.
Capability inventory: The agent has the capability to execute validation scripts (npm test, pytest, make) and modify repository state via gh pr merge.
Sanitization: The skill does not implement sanitization or filtering of the comment content before the agent processes it to address feedback.

workflow-pr