The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted input from GitHub issues, creating an attack surface for indirect prompt injection. * Ingestion points: GitHub issue bodies and comments are retrieved via the gh issue view command in SKILL.md. * Boundary markers: Absent. The agent processes the raw output of the command without explicit delimiters. * Capability inventory: The skill can modify repository state, such as labeling, closing, or transferring issues, using gh issue commands in SKILL.md. * Sanitization: No programmatic sanitization is applied to the retrieved content, although Step 4 explicitly instructs the agent to perform manual verification and reproduction.

triage-issue