ai-podcast-creation

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill includes functionality to generate podcast scripts from user-provided documents, creating an indirect prompt injection surface where document content could potentially inf luence the LLM. This is an intended feature of the skill.
  • In gestion points: Document content is processed by the openrouter/claude-sonnet-45 application.
  • Boundary markers: None; the prompt doe s not use delimiters to separate instructions from untrusted data.
  • Capability inventory: The skill uses belt app run and can write output to files.
  • Sanitization: None; the document text is not sanitized before being passed to the model.
  • [EXTERNAL_DOW NLOAD S]: The skill reference s documentation and additional skill s from the vendor's of f icial GitHub repositor y (in ference-sh). The se are legitimate vendor resource s.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 10:14 PM