podcast-production
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external transcript data which could contain malicious instructions meant to influence the agent's behavior during the creation of cold opens, titles, or blog posts.\n
- Ingestion points: The primary input is
[Guest]_Source_Material.md, which contains raw transcript data provided to the skill.\n - Boundary markers: The instructions lack specific delimiters or "ignore embedded instructions" warnings when processing the transcript data, relying only on "Verbatim Only" constraints for quotes.\n
- Capability inventory: The skill possesses file-writing capabilities and invokes multiple sub-skills (e.g.,
youtube-title-creator,cold-open-creator,brand-identity-wizard) which increases the impact if an injection occurs.\n - Sanitization: No specific technical sanitization or validation of the transcript content is performed beyond formatting and cleaning for readability.
Audit Metadata