The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill features a large attack surface for indirect prompt injection due to its core functionality of aggregating information from external platforms to guide document creation.
Ingestion points: The workflow explicitly prompts the user to provide context from messaging platforms (Slack, Microsoft Teams), shared document storage (Google Drive, SharePoint), and shared document links, as well as text and images pasted directly into the conversation.
Boundary markers: The instructions do not define any delimiters or provide specific guidance for the agent to disregard instructions that might be embedded within the external documents or messaging history it processes.
Capability inventory: The skill possesses the ability to create and modify local files using tools like create_file and str_replace, and it can invoke sub-agents to test the generated documentation.
Sanitization: There is no mechanism described for validating, filtering, or sanitizing the content retrieved from external integrations before it is incorporated into the agent's context for drafting and refinement.

doc-coauthoring