Documentation
Open-source security platform for AI agents. Pick a tool to get started.
Quick start. Start with the CLI. One command to scan, protect, and monitor.
npx opena2a-cli review
Scan & Fix
Find vulnerabilities, test defenses, and protect credentials
OpenA2A CLI
Unified security platform. Scan, protect, monitor, and detect Shadow AI agents and MCP servers.
HackMyAgent
Security scanner, red-team toolkit, OASB benchmark, and runtime protection. 318 static + 29 semantic + 164 adversarial checks. Auto-fix with rollback.
Secretless
Credential protection for AI tools. Prevents key leakage in LLM contexts.
NanoMind
On-device semantic classifier embedded in HackMyAgent. Sorts agent artifacts into ten attack classes locally, with no cloud calls.
Identity & Trust
Agent identity, browser protection, and supply chain trust
AIM
Agent Identity Management. Ed25519 identity, trust scoring, MCP security, ABOM.
Browser Guard
Chrome extension for detecting and monitoring browser-based AI agents.
Trust Data
Query the trust level and score the OpenA2A Registry holds for an AI package, run security scans, and share scan summaries with the Registry.
Train
Security training environments
Use Cases
Step-by-step workflows for your role
Interactive Demos
See the tools in action. Run guided security scenarios in your browser.
Pick the right tool for your use case
Follow the getting-started guide.