The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from the repository's environment to generate PR metadata.
Ingestion points: The skill reads commit messages via git log, branch names via git branch, and the contents of .github/pull_request_template.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat commit messages or branch names as untrusted data that should not influence its logic.
Capability inventory: The skill has the capability to execute shell commands (git push, gh pr create) and perform network operations to GitHub.
Sanitization: There is no mention of sanitizing or escaping the content extracted from commits before it is used in the gh pr create command or populated into the PR template.
[COMMAND_EXECUTION]: The skill executes multiple shell commands to manage Git operations and interface with the GitHub CLI.
The skill uses git log, git diff, and git remote to gather context.
The skill uses gh pr create and gh pr view to manage remote PRs.
The skill includes safety checks such as gh auth status and git status to ensure a clean state before execution.

create-pull-request