funding-skills
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and provide code for fetching research data from official and reputable public APIs including the NIH RePORTER, NSF Award Search, Figshare, and Zenodo. These represent legitimate, well-known research infrastructure services.
- [COMMAND_EXECUTION]: Example curl commands are included in the documentation to assist users in verifying API connectivity and testing search queries against official endpoints.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external research data (e.g., award abstracts and project titles) from public APIs. 1. Ingestion points: JSON responses from NIH, NSF, Figshare, and Zenodo API endpoints. 2. Boundary markers: Not explicitly implemented in the provided illustrative templates. 3. Capability inventory: The skill logic focuses on network requests and data presentation. 4. Sanitization: Not addressed in the high-level educational snippets. This is considered a safe surface as it is integral to the primary research function of the skill.
Audit Metadata