The Agent Skills Directory

[SAFE]: The skill utilizes standard GitHub CLI (gh) and git commands for repository and project management. All actions that modify state, such as creating issues or adding items to project boards, are gated by mandatory user confirmation steps.
[SAFE]: The skill implements defensive measures against indirect prompt injection by instructing the agent to treat existing GitHub issue content (titles, bodies, comments) as untrusted. It explicitly forbids following any instructions embedded within that content, using it only for duplicate detection.
[SAFE]: No evidence of unauthorized data exfiltration, hardcoded credentials, or obfuscated code was found. The skill operates within the boundaries of the targeted GitHub repository and project environment.

feature-intake