Skip to main content
The security_audit prompt instructs the AI agent to perform a systematic security review across all managed wallets and the system’s overall configuration.

​
When It’s Used

Invoked by operators who want a compliance-style review of their deployment — typically on a regular schedule or after a security incident.

​
Prompt Template

System Prompt: “You are an expert Solana smart contract security auditor. Perform a comprehensive security audit of this Agentic Wallet deployment. Review all wallet policies, analyze the audit log for violations and anomalies, check spending patterns, and provide a security posture assessment with a 1-100 risk score.”

​
What the Agent Does

  1. Reads system://status and system://config for infrastructure health
  2. Lists all wallets via list_wallets and checks their policies
  3. Reviews audit://logs for policy violations and blocked transactions
  4. Analyzes spending patterns for anomalies (sudden spikes, unusual recipients)
  5. Reports:
    • Number of policy violations detected
    • Unusual transaction patterns
    • Blocked transaction summary
    • Overall security posture score (1-100)
    • Recommended remediation actions

​
Benefits

  • Automated compliance: Run security reviews without manual log analysis
  • Pattern detection: LLMs can identify subtle anomalies humans might miss in raw log data
  • Actionable output: The audit produces specific, prioritized recommendations