doc-coauthoring

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes file system tools (create_file, str_replace) and sub-agent invocation to manage document drafting and testing. These are used as intended for document management and validation within the agent's environment.
  • [REMOTE_CODE_EXECUTION]: The skill mentions external integrations (Slack, Google Drive, etc.) via MCP servers or connectors. Access to these resources is mediated by the platform's permission model and user confirmation, following safe practice for context retrieval.
  • [DATA_EXFILTRATION]: While the skill reads external context from shared documents and channels, it does so to process the data locally for the user's drafting task. No evidence of sending sensitive data to unauthorized external endpoints was found.
  • [PROMPT_INJECTION]: The workflow includes explicit instructions for handling external data (Reader Testing), utilizing fresh agent instances to avoid context bleed and prevent obedience to potential instructions embedded in the processed documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 05:03 AM
Security Audit — agent-trust-hub — doc-coauthoring