podcast-episode-producer

Pass

Audited by Gen Agent Trust Hub on Mar 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill operates as a legitimate automation tool for podcast publishing. It processes user-provided transcripts and follows pre-configured brand and writing standards stored in the references directory. All external domains listed in the configuration belong to the skill's author or represent standard platform links.- [PROMPT_INJECTION]: Evaluation of Indirect Prompt Injection surface (Category 8):
  • Ingestion points: The skill ingests untrusted external text in the form of podcast transcripts provided by the user in SKILL.md.
  • Boundary markers: There are no explicit delimiters or instructions used to separate the transcript content from the skill's operational instructions.
  • Capability inventory: The skill generates downstream prompts for external AI tools (NotebookLM) and image generation briefs.
  • Sanitization: The input text is summarized and formatted without active sanitization or character filtering.
  • Context: This identifies a potential vulnerability surface common to all LLM-based summary tools; no malicious intent or exploitation attempts were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 28, 2026, 10:15 PM