doc-coauthoring

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing untrusted data from multiple sources.
  • Ingestion points: The workflow ingests content from external sources including local files provided by the user, shared document platforms (Google Drive, SharePoint), and messaging channels (Slack, Teams).
  • Boundary markers: The skill does not define explicit delimiters or 'ignore' instructions to isolate processed context from its internal workflow logic.
  • Capability inventory: The agent uses create_file to generate document scaffolds and str_replace to perform iterative edits on the filesystem.
  • Sanitization: No explicit sanitization, validation, or filtering of external content is specified before the data is integrated into the prompt context.
  • [EXTERNAL_DOWNLOADS]: Fetches document context and templates from external integrations and links. These operations utilize well-known services (e.g., Google Drive, Slack) and typically require user-enabled connectors or explicit confirmation.
  • [COMMAND_EXECUTION]: Uses local file system tools to create and modify documents. The operations (create_file, str_replace) are consistent with the skill's primary purpose of document co-authoring and do not involve arbitrary command execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 10:49 PM
Security Audit — agent-trust-hub — doc-coauthoring