The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified. The skill performs its stated functions using standard, trusted developer tools.
[COMMAND_EXECUTION]: The skill utilizes bash commands for git and the GitHub CLI (gh) to query repository state and manage Pull Requests.
[INDIRECT_PROMPT_INJECTION]: The skill processes commit messages to generate Pull Request metadata, which represents a potential ingestion surface for untrusted data.
Ingestion points: Commit history retrieved via git log commands in SKILL.md.
Boundary markers: Not present in the command templates for PR creation.
Capability inventory: The skill has the capability to write to the GitHub API via gh pr create (SKILL.md).
Sanitization: The skill uses raw commit messages directly for PR titles and bodies without sanitization or validation.

git-pr-create