tern
Warn
Audited by Snyk on Oct 21, 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 agent reads output and messages from external coding agents running in terminal panes using
tern captureandagentmsg.py read, which can ingest outsider-authored free text from untrusted agent outputs and incoming messages.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata