The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git, gh (GitHub CLI), and bun (JavaScript runtime) to perform repository operations, run builds, and manage pull requests. Evidence: Shell commands like git diff, bun run build, and gh pr create are defined in SKILL.md to automate the development workflow.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from the repository, such as git diffs and pull request templates. 1. Ingestion points: The skill reads git diff output, .github/pull_request_template.md content, and existing PR metadata via gh pr view into the agent's context. 2. Boundary markers: There are no explicit delimiters or specific instructions for the agent to ignore potentially malicious instructions embedded in the code diffs or templates. 3. Capability inventory: The skill possesses the ability to execute shell commands, push to remote repositories, and create or edit pull requests. 4. Sanitization: The skill implements a security best practice by using a single-quoted heredoc ('EOF') in the gh pr create and gh pr edit commands, which prevents the shell from executing any potentially malicious content generated within the PR body.

open-pr2