The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill uses standard tools for their intended purpose.
[COMMAND_EXECUTION]: The skill uses git and gh for repository management. All commands follow established workflows for pull request creation and branch rebasing.
[DATA_EXFILTRATION]: Communication with external services is limited to GitHub via the official gh CLI tool. No access to sensitive local credentials or configuration files was detected.
[PROMPT_INJECTION]: The skill processes external data from repository logs and pull request views. 1. Ingestion points: SKILL.md (lines referencing gh pr view and git log output). 2. Boundary markers: Absent. 3. Capability inventory: SKILL.md (uses git push and gh pr edit to modify remote state). 4. Sanitization: Absent. This data ingestion is a standard requirement for pull request management and does not constitute a malicious attempt to override agent behavior.

manage-pr