The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands using git and the GitHub CLI (gh) to read repository state, manage local branches, and interface with the GitHub API for pull request operations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local repository and remote GitHub templates to generate pull request titles and descriptions.
Ingestion points: Processes data from git diff, git log, and repository-level or organization-level PULL_REQUEST_TEMPLATE.md files (SKILL.md).
Boundary markers: No specific delimiters or safety instructions are defined to separate code-derived content from agent instructions during PR body assembly.
Capability inventory: The skill has the ability to perform state-changing actions, including git push to remote origins and gh pr create / gh pr edit to modify pull requests (SKILL.md).
Sanitization: The skill mitigates risks by requiring explicit user approval (Step 8) before executing any commands that modify the repository or GitHub state.
[DATA_EXPOSURE_AND_EXFILTRATION]: Fetches pull request templates from remote GitHub repositories using the GitHub API (gh api). This operation targets official GitHub infrastructure to resolve organization-level configurations.

pr