crypt-librarian
Warn
Audited by Snyk on Nov 6, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The required workflow uses a script (
scripts/flexible_discovery.py) that queries the public Reddit JSON API (search.json) to ingest arbitrary, outsider-authored free text into the agent's execution path without prior selection.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata