doc-coauthoring
Pass
Audited by Gen Agent Trust Hub on Apr 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes file system operations including
create_fileto initialize documentation scaffolds andstr_replaceto apply iterative refinements to specific sections of the document. - [EXTERNAL_DOWNLOADS]: The workflow facilitates fetching information from well-known services such as Slack, Google Drive, and SharePoint via platform integrations or MCP servers to build context for the drafting process.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it ingests and processes data from external documents and messaging platforms.
- Ingestion points: External context is gathered from chat threads, linked documents, and user-provided files as described throughout the Stage 1 workflow in
SKILL.md. - Boundary markers: The instructions lack explicit delimiting markers or warnings for the agent to ignore instructions embedded within the ingested external content.
- Capability inventory: The agent has the capability to write and modify the local file system using
create_fileandstr_replacebased on the processed context. - Sanitization: There are no documented mechanisms for sanitizing or validating external content before it is processed by the agent to influence document drafting.
Audit Metadata