octocode-research
Warn
Audited by Snyk on Feb 15, 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). This skill explicitly performs external GitHub/npm research ("Handles both local codebase analysis ... and external GitHub/npm research using Octocode tools") and requires fetching and interpreting public repository pages/URLs (see Phase descriptions and required GitHub URL references), exposing the agent to untrusted, user-generated third‑party content that could carry indirect prompt injection.
Audit Metadata