alterlab-nmc-media-theory
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines an autonomous workflow that processes untrusted data from external sources and user-provided files, which can be exploited via indirect prompt injection.
- Ingestion points: The workflow instructs the agent to "search the web for academic databases, journal articles, theory frameworks" and "read existing project files (draft papers, research proposals, reading lists, course syllabi)" as primary inputs.
- Boundary markers: The instructions do not define clear boundaries or provide guidance to the agent to disregard instructions potentially embedded within the retrieved academic papers or user documents.
- Capability inventory: The skill possesses the capability to perform web searches, read local files, and write new Markdown files to the project directory.
- Sanitization: There are no specified sanitization or validation steps for content ingested from external web sources before it is used to generate deliverables.
Audit Metadata