wikidata-search
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from Wikidata, which is a public and community-editable knowledge base. This creates a potential surface for indirect prompt injection attacks.\n
- Ingestion points: Data is retrieved from external endpoints (wikidata.org and wmcloud.org) via the WikidataAPI class in scripts/wikidata_api.py.\n
- Boundary markers: The instructions do not define clear boundary markers or provide the agent with guidance on how to handle potentially untrusted content within search results.\n
- Capability inventory: The skill has network access capabilities through the urllib.request library and can read local environment variables for API authentication.\n
- Sanitization: The retrieved data is parsed as JSON but does not undergo specific sanitization to filter out malicious instructions or patterns.
Audit Metadata