The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill is vulnerable to indirect prompt injection because its core function involves reading untrusted external data (source code and existing documentation).
Ingestion points: The skill explicitly instructs the agent to 'Read the relevant source code' and 'Review existing related documentation' in SKILL.md.
Boundary markers: There are no instructions provided to the agent to treat the ingested code strictly as data or to ignore embedded instructions (e.g., instructions hidden in code comments).
Capability inventory: The skill description states it is used to 'generate, update, or refactor' documentation, implying file-write or file-modification permissions which elevates the risk of successful injection.
Sanitization: No sanitization or validation steps are defined for the content being read before it is processed by the agent.

docs-out