nlweb-schema-org-grounding
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill structures and processes data from external websites, which serves as a potential vector for indirect instructions.
- Ingestion points: The skill instructions involve fetching and auditing Schema.org JSON-LD from external sites via tools like
WebFetch. - Boundary markers: There are no specific delimiters or instructions provided to the agent to distinguish between its operational instructions and the potentially malicious data content within the JSON-LD fields.
- Capability inventory: The skill has access to tools such as
Bash,Write,Edit, andWebFetchas listed in the YAML metadata. - Sanitization: The skill emphasizes structural validation (e.g., using Google's Rich Results Test) but does not address sanitization of the textual content within JSON-LD fields that the agent may process.
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch configuration and reference documents from a public GitHub repository.
- Evidence: References to
https://github.com/nlweb-ai/NLWeb/blob/main/config/sites.xmlandhttps://github.com/nlweb-ai/NLWeb/blob/main/docs/nlweb-prompts.mdin the implementation guidance.
Audit Metadata