openaire-search
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches research products, projects, and organization metadata from the OpenAIRE Graph API (api.openaire.eu). This external content is ingested into the agent's context, creating a potential surface for indirect prompt injection if the retrieved metadata contains malicious instructions.
- Ingestion points: Data is fetched in scripts/openaire_api.py via the OpenAIREAPI class methods.
- Boundary markers: The instructions do not specify the use of delimiters or isolation markers for the external data returned to the agent.
- Capability inventory: The skill utilizes network requests to the OpenAIRE API; the agent typically has access to shell and file system tools.
- Sanitization: The script includes a _clean function that removes HTML and JATS tags from the retrieved text and truncates descriptions to a maximum of 1000 characters.
Audit Metadata