The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from VCS issues and pull requests, creating a surface for indirect prompt injection.
Ingestion points: SKILL.md (Workflow Step 3 mentions reading original descriptions and all comments).
Boundary markers: Absent. The instructions do not specify delimiters or instructions to ignore commands within the ingested text.
Capability inventory: SKILL.md leverages gh and glab CLI tools for repository cloning and posting comments.
Sanitization: Absent. There is no evidence of filtering or validation for the content read from issues.
[COMMAND_EXECUTION]: The skill uses specific command-line tools for its operations.
Evidence: SKILL.md (Workflow Steps 2 and 5) specifies the use of gh repo clone, gh issue/pr comment, and glab issue/mr note.

respond-github