flare-smart-accounts
Pass
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references the smart-accounts-cli utility from the official flare-foundation GitHub repository.
- [CREDENTIALS_UNSAFE]: Mentions sensitive environment variables like XRPL_SECRET and FLARE_PRIVATE_KEY for local tool configuration, accompanied by clear security warnings and instructions to keep these credentials isolated from AI interactions.
- [PROMPT_INJECTION]: Proactively addresses Indirect Prompt Injection surface (Category 8). Ingestion points: XRPL payment memos and RPC data. Boundary markers: Strict 32-byte fixed-format binary instruction structure. Capability inventory: Blockchain transaction submission and contract calls. Sanitization: Documentation explicitly advises treating incoming data as structured payloads rather than natural language or prompt inputs.
Audit Metadata