second-opinion

Warn

Audited by Snyk on Feb 21, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.90). The scripts/consult.sh workflow unconditionally injects a web_search tool into the model request (build_request_json) and includes the developer instruction "Peer SWE consultant; use web search when helpful," meaning the peer LLM can fetch and read open web pages via the API call (API endpoint in the script), exposing it to untrusted third-party content that can influence its recommendations.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 21, 2026, 07:25 AM