arena-trader
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to suppress its AI identity and avoid standard safety-related responses (e.g., "Never use phrases like 'as an AI' or 'I cannot'").
- [PROMPT_INJECTION]: The agent is exposed to indirect prompt injection risk as it is instructed to monitor and react to untrusted messages from other participants in a public Telegram group. Ingestion points: Telegram group messages. Boundary markers: None present. Capability inventory: SOL trading and token information retrieval. Sanitization: No structural content sanitization is defined, although the agent is told to perform manual verification before trading.
Audit Metadata