twinmind-common-errors

Warn

Audited by Snyk on Apr 4, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly directs checking and fetching public endpoints (e.g., curl https://status.twinmind.com in Quick Diagnostic Commands and "Search community forum: https://community.twinmind.com" in the Escalation Path), meaning the workflow expects the agent/user to ingest and act on untrusted, user-generated third‑party forum/status content that could influence debugging decisions.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 4, 2026, 07:26 PM
Issues
1