The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it retrieves content from external, potentially attacker-controlled Google Documents.
Ingestion points: The read_document tool in handler.ts fetches full text content from the Google Docs API and returns it to the agent.
Boundary markers: The tool definitions in tools.ts and the execution logic in handler.ts do not implement boundary markers or instructions to treat the document content as untrusted data.
Capability inventory: The skill provides tools for document listing, creation, and modification (list_documents, create_document, append_to_document), which could be abused if the agent follows instructions found within a read document.
Sanitization: There is no evidence of sanitization, filtering, or escaping of the content retrieved from the Docs API before it is processed by the AI.

google-docs