grounded-citations

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implementation follows security best practices for citation management. It uses a local JSON ledger to store source mappings and verifies evidence through verbatim string matching, which acts as a defense against model hallucinations and misinformation.
  • [SAFE]: The provided scripts (sources.py and _hermes_home.py) rely exclusively on the Python standard library, avoiding risks associated with third-party dependencies or external downloads.
  • [SAFE]: File operations are restricted to the local filesystem for managing the citation ledger (by default in ~/.hermes) and reading files provided as arguments by the agent, with no signs of malicious exfiltration or persistence mechanisms.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 08:54 AM
Security Audit — agent-trust-hub — grounded-citations