The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill features a vulnerability surface for indirect prompt injection as it processes untrusted data from external GitHub resources.
Ingestion points: The agent retrieves external content using gh pr view, gh pr diff, and gh pr review as outlined in SKILL.md and reference/pr-workflow.md.
Boundary markers: The instructions do not establish explicit boundary markers or delimiters to differentiate between system instructions and data ingested from PRs or issues.
Capability inventory: The skill possesses the capability to execute complex git and gh commands, enabling it to modify local files, create branches, and push changes to remote repositories.
Sanitization: There is no evidence of sanitization, filtering, or validation processes for the content retrieved from external GitHub descriptions or comments before it is added to the agent's context.

github-workflow