The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external issue trackers and pull requests, which creates a potential surface for indirect prompt injection if the source data contains malicious instructions intended to influence the agent.
Ingestion points: Content is retrieved from ticket or pull request links and numbers as specified in SKILL.md.
Boundary markers: Technical delimiters are absent in SKILL.md, although the instructions tell the agent to distinguish between historical context, current evidence, and proposed changes.
Capability inventory: The skill triggers mattpocock-skills:diagnosing-bugs to perform verification tests and skills:grill-minimal to discuss implementation changes, as described in SKILL.md.
Sanitization: No specific methods for sanitizing or escaping retrieved ticket content are mentioned in the skill instructions.

revisit