ai-podcast-creation
Fail
Audited by Snyk on Feb 16, 2026
Risk Level: CRITICAL
Full Analysis
CRITICAL E005: Suspicious download URL detected in skill instructions.
- Suspicious download URL detected (high risk: 0.80). These URLs are hosted on an unvetted third‑party domain that advertises a curl | sh installer (https://cli.inference.sh), and piping a remote shell script from an unknown source is a high‑risk pattern even if the image and site themselves look harmless.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests arbitrary user-supplied or external documents into an LLM (see "NotebookLM-Style Content" where openrouter/claude-sonnet-45 is invoked with "Document: ") and also accepts arbitrary audio URLs for media merging, so the agent would read and act on untrusted third‑party content.
Audit Metadata