duo-expert-content-engine
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of Markdown documentation and strategic guidelines. There are no executable scripts (.py, .js, .sh), configuration files that trigger shell commands, or installation manifests (package.json, requirements.txt) included in the package.- [INDIRECT_PROMPT_INJECTION]: The documentation describes a business methodology for ingesting external, untrusted user data (reader questions, corrections, and crowdsourced tactics) into a publishing workflow. While the skill advocates for this pattern, it does not provide an automated tool to perform the ingestion.
- Ingestion points: Described in
references/reader-questions-as-pre-validated-demand.mdandreferences/promote-corrections-to-their-own-instalment.mdas inbound email and comment systems. - Boundary markers: The strategy suggests the use of "pun handles" and "illustrated letters" which function as human-in-the-loop delimiters.
- Capability inventory: The skill contains no code execution, file writing, or network capabilities.
- Sanitization: The strategy explicitly recommends an editorial selection and anonymization process for all ingested content.
Audit Metadata