know-your-unknowns

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform codebase analysis using shell commands like git log and grep to find historical failures, usage patterns, and existing conventions.
  • [EXTERNAL_DOWNLOADS]: Technique 07 ("Point at a reference") involves the agent reading and analyzing external source code from repositories or gists to create semantic comparisons for porting code.
  • [SAFE]: The instructions include a specific security rule requiring all generated HTML artifacts to be self-contained and offline-capable, explicitly forbidding any external network calls, CDN usage, or remote resource loading (e.g., zero external requests, no CDN).
  • [SAFE]: The interaction model uses navigator.clipboard.writeText to help the user copy data back into the chat interface, maintaining clear user oversight over the data being shared with the model.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 01:49 PM
Security Audit — agent-trust-hub — know-your-unknowns