hive-nft-research
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted metadata from external NFT collections, representing a surface for indirect prompt injection attacks.
- Ingestion points: Data enters the agent's context through
invoke_api_endpoint(SKILL.md) when fetching metadata, traits, and provenance data from external blockchains and marketplaces. - Boundary markers: The skill lacks explicit delimiters or specific instructions for the agent to ignore any natural language commands that might be embedded within the retrieved NFT metadata strings.
- Capability inventory: The skill has the ability to search tools, retrieve API schemas, and invoke endpoints via the
nft_researchtoolset. - Sanitization: While the
validate_task_resulttool is mentioned for structural verification, there is no evidence of filtering or sanitizing natural language content within the raw metadata to prevent instruction injection.
Audit Metadata