The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external GitHub users, which is a known vector for indirect prompt injection attacks.
Ingestion points: External data is fetched via gh issue view, gh pr list, and gh pr view in SKILL.md.
Boundary markers: Absent. There are no specific instructions to use delimiters to isolate untrusted content.
Capability inventory: The skill allows the agent to modify repository metadata and post comments using gh issue edit and gh issue comment.
Sanitization: Present. The skill instructions require using the --body-file flag for comments to prevent shell command injection from user-provided text.
[COMMAND_EXECUTION]: The skill relies on shell-based gh CLI commands to interact with the GitHub platform, enabling the agent to read from and write to the repository.

review-issue