ai-podcast-creation

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface in the 'NotebookLM-Style Content' workflow. Ingestion points: External document content is interpolated directly into a prompt for the Claude LLM via 'SKILL.md'. Boundary markers: The prompt template lacks delimiters or specific instructions to ignore embedded commands within the ingested document. Capability inventory: The skill utilizes the 'belt' CLI to execute various media processing and text generation tasks, which could be influenced by malicious instructions in the input data. Sanitization: No sanitization or validation of the input document is performed before processing.
  • [EXTERNAL_DOWNLOADS]: The skill documentation references a remote installation script for the 'belt' CLI tool hosted on the vendor's official GitHub repository at 'raw.githubusercontent.com/inference-sh/skills/'.
  • [COMMAND_EXECUTION]: The skill uses the 'belt' CLI command to orchestrate remote services for audio synthesis, music generation, and file merging.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 07:15 PM
Security Audit — agent-trust-hub — ai-podcast-creation