The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources, which can be used to influence the agent's behavior via hidden instructions.
Ingestion points: SKILL.md specifies that the agent should investigate primary sources including official documentation, source code, and first-party APIs.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to ignore or isolate instructions potentially contained within the processed external data.
Capability inventory: The agent has the capability to write files to the repository's filesystem as specified in SKILL.md (item 3).
Sanitization: The skill does not define any mechanisms for sanitizing, filtering, or validating the content retrieved from external sources before processing it.

research