hive-defi-research
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external DeFi metrics (TVL, fees, yields) from API providers, creating a potential indirect prompt injection surface. The skill mitigates this through a mandatory 'Evidence receipt' system, structured report templates, and a
validate_task_resulttool to verify the output's structure and provenance. Ingestion points include data frominvoke_api_endpoint(SKILL.md). The capability inventory is limited to authenticated toolset calls with no file-system write or subprocess execution capabilities. Boundary markers are provided by the structured report template. - [SAFE]: Analysis of the instructions and configuration files revealed no evidence of prompt injection, data exfiltration, or obfuscation. All identified toolsets and packages are within the developer's official hive-intel ecosystem.
Audit Metadata