The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted data from external sources and has file-writing capabilities.\n
Ingestion points: Content is retrieved from messaging apps (Slack, Teams) and shared document storage (Google Drive, SharePoint) during the context gathering stage (SKILL.md).\n
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands in the fetched data.\n
Capability inventory: The agent uses create_file to establish document structures and str_replace to refine drafted sections.\n
Sanitization: No sanitization or validation of external content is specified before processing.\n- [EXTERNAL_DOWNLOADS]: The skill references integrations with well-known productivity services including Google Drive, SharePoint, and Slack to pull context. These references are part of the intended document workflow and are documented neutrally as they target established platforms.\n- [COMMAND_EXECUTION]: The skill instructs the agent to use create_file and str_replace tools to generate and update document artifacts. These commands are used solely for managing the document draft within the user's workspace.

doc-coauthoring