The Agent Skills Directory

[SAFE]: The skill implements a standard development workflow using expected command-line tools such as git and gh.
[SAFE]: All identified operations, including checking repository status, gathering commit logs, diffing changes, pushing branches, and creating pull requests, are consistent with the skill's stated purpose.
[SAFE]: No evidence of prompt injection, obfuscation, unauthorized data exfiltration, or malicious persistence mechanisms was detected within the instructions or command sequences.
[SAFE]: While the skill processes code diffs which technically constitute untrusted data (Indirect Prompt Injection surface), this behavior is the primary intended function for generating PR summaries, and the resulting output is constrained to PR metadata handled by the GitHub CLI.

create-pr