The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources such as GitHub issue bodies, pull request comments, and npm registry data.
Ingestion points: SKILL.md instructions for reading content via gh issue view, gh pr view, gh api, and npm view.
Boundary markers: The skill lacks explicit instructions for the agent to use delimiters or ignore embedded instructions when processing external data.
Capability inventory: The skill utilizes Bash to execute commands for repository management and metadata retrieval.
Sanitization: No explicit sanitization or filtering of external content is described before interpolation into the agent's context.

github-ops