The Agent Skills Directory

[SAFE]: No malicious activities, obfuscation, or unauthorized data access patterns were identified. The skill's primary function is to read local git data and create a Pull Request via the standard GitHub CLI tool.
[PROMPT_INJECTION]: The skill was evaluated for indirect prompt injection as it ingests untrusted data from git history. * Ingestion points: SKILL.md (Step 1: commit messages and diffs). * Boundary markers: Not specified. * Capability inventory: gh pr create (GitHub CLI) for creating remote resources. * Sanitization: None mentioned, however, Step 3 requires explicit user confirmation via AskUserQuestion before the PR is created, which serves as an effective mitigation for automated injection attacks.

create-pr