goat
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill documents how to build autonomous onchain agents with extensive capabilities, creating a surface for indirect prompt injection where malicious external data could influence the agent's financial actions.
- Ingestion points: User-provided prompts and onchain data processed by the agent in
SKILL.md. - Boundary markers: The provided implementation examples do not specify explicit boundary markers (e.g., delimiters) to isolate instructions from data.
- Capability inventory: Full wallet access, including
sendTransaction,sendETH, andsendSOL, plus arbitrary smart contract calls and DEX swaps across 30+ chains as documented in the plugin and wallet provider sections. - Sanitization: The documentation explicitly includes a "Best Practices" section advising developers to guard against prompt injection, validate inputs, set spending limits, and require human confirmation.
Audit Metadata