Get Your Agent Certified

Cryptographic identity + public trust score for your AI agent. Under 5 minutes. Works on any platform.

Step 1 — Install

OpenClaw / Hermes Agent

clawhub install agentshield

All other platforms (LangChain, CrewAI, AutoGen, n8n, …)

pip install agentshield-audit

Step 2 — Run your first audit

# Auto-detect (OpenClaw, Hermes, n8n)
python initiate_audit.py --auto

# Manual (LangChain, CrewAI, custom)
python initiate_audit.py --name "MyAgent" --platform langchain

This runs 77 security tests locally (~30 seconds). Only pass/fail scores are sent to the API — no prompts, no responses, no private data. → Full privacy details

Step 3 — View your certificate

python show_certificate.py

Your Ed25519 private key stays local. The signed certificate is stored in ~/.agentshield/certificate.json

Step 4 — Public verification

https://agentshield.live/verify/<your_agent_id>

Anyone can verify your agent's trust score and certificate — no account needed. → Open Verify page

Platform Compatibility

Platform Status Flag
OpenClaw ✅ Full support --auto
Hermes Agent ✅ Full support --auto
n8n ✅ Auto-detected --auto
LangChain / LangGraph ✅ Manual --name "X" --platform langchain
CrewAI ✅ Manual --name "X" --platform crewai
AutoGen / Custom ✅ Manual --name "X" --platform cli

Early Adopter Program

We're looking for the first 25 serious agent builders. Install, run 1–2 audits, give honest feedback. You get a public Early Adopter Badge + priority feature access.

🚀 Join as Early Adopter →
Full Documentation  ·  GitHub  ·  Verify an Agent