A2A Demos

v1.0.0Securitystable

Demo agents showcasing CapiscIO Agent Guard and MCP Guard — trust badges, identity verification, and tool-level authorization for A2A and MCP protocols

a2aa2a-protocolagent-to-agentai-agentsai-governance
Share:
22
Stars
0
Downloads
0
Weekly
0/5

What is A2A Demos?

A2A Demos is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to demo agents showcasing capiscio agent guard and mcp guard — trust badges, identity verification, and tool-level authorization for a2a and mcp protocols

Demo agents showcasing CapiscIO Agent Guard and MCP Guard — trust badges, identity verification, and tool-level authorization for A2A and MCP protocols

This server falls under the Security category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Demo agents showcasing CapiscIO Agent Guard and MCP Guard —

Use Cases

Agent-to-agent security
Identity verification
Tool-level authorization
capiscio

Maintainer

LicenseMIT
Languagepython
Versionv1.0.0
UpdatedMay 16, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx a2a-demos

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use A2A Demos

CapiscIO A2A Demos is a collection of working demo agents that showcase CapiscIO's trust infrastructure for AI — described as 'Let's Encrypt for AI'. The demos cover cryptographic identity verification via Decentralized Identifiers (DIDs), trust badge issuance, per-tool enforcement policies (min_trust_level), runtime policy changes without code redeployment, and multi-framework agent integration using LangChain, CrewAI, and LangGraph. They are designed to help developers understand how to add identity, authorization, and policy enforcement to MCP servers and A2A agents.

Prerequisites

  • Python 3.11 or higher
  • A free CapiscIO account at app.capisc.io with an API key (Dashboard → Settings → API Keys)
  • OpenAI API key (required only for the multi-agent integration demos, not the concept demos)
  • An MCP client such as Claude Desktop or Claude Code if integrating via MCP
1

Clone the repository

Clone the a2a-demos repository from GitHub to get all demo scripts and setup files.

git clone https://github.com/capiscio/a2a-demos
cd a2a-demos
2

Set up the Enforcement Demo (concept demo, no LLM needed)

Navigate to the enforcement-demo directory and run the setup script. It creates a virtual environment, installs dependencies, and downloads a required binary (~15 MB). Then edit the generated .env file to add your CAPISCIO_API_KEY.

cd enforcement-demo
./setup.sh
# Edit .env: add CAPISCIO_API_KEY=sk_live_your_key_here
3

Run the Enforcement Demo

Activate the virtual environment and run the demo. It verifies the MCP server's cryptographic identity and runs four enforcement scenarios showing which agents are allowed or denied based on their trust badge.

source .venv/bin/activate
python run_demo.py
4

Set up the Policy Demo (optional, no LLM needed)

Navigate to policy-demo, run setup, edit .env, then create the three runtime policies in the CapiscIO registry before running the interactive demo that pauses between phases.

cd ../policy-demo
./setup.sh
# Edit .env: add CAPISCIO_API_KEY
source .venv/bin/activate
python scripts/setup_policies.py
python run_demo.py
5

Set up the Multi-Agent Integration Demo (requires OpenAI key)

Navigate to multi-agent-demo and run setup. This creates per-agent virtual environments for the LangChain, CrewAI, and LangGraph agents and installs a shared CapiscIO module.

cd ../multi-agent-demo
./setup.sh
6

Configure environment for multi-agent demo

Edit the generated .env file with both your OpenAI and CapiscIO credentials.

# Edit multi-agent-demo/.env:
# OPENAI_API_KEY=sk-your-openai-key
# OPENAI_MODEL=gpt-4o-mini
# CAPISCIO_API_KEY=sk_live_your_api_key_here
# CAPISCIO_SERVER_URL=https://registry.capisc.io
# SECURITY_MODE=ca
7

Run multi-agent demo

Launch each agent in its own terminal or use the provided tmux script to start all three agents (LangChain on port 8001, CrewAI on port 8002, LangGraph on port 8003) then run the demo driver.

# Option A: all at once
./run-agents.sh

# Option B: individually
# Terminal 1
cd agents/langchain-agent && source .venv/bin/activate && python main.py --serve
# Terminal 2
cd agents/crewai-agent && source .venv/bin/activate && python main.py --serve
# Terminal 3
cd agents/langgraph-agent && source .venv/bin/activate && python main.py --serve

# Demo driver
source agents/langchain-agent/.venv/bin/activate
python run_demo.py --discover

A2A Demos Examples

Client configuration

These are concept and integration demos rather than a standalone MCP server package. The enforcement demo runs as a local Python process. Example MCP server registration for an MCP-enabled CapiscIO-secured server:

{
  "mcpServers": {
    "capiscio-demo": {
      "command": "python",
      "args": ["run_demo.py"],
      "cwd": "/path/to/a2a-demos/enforcement-demo",
      "env": {
        "CAPISCIO_API_KEY": "sk_live_your_key_here",
        "CAPISCIO_SERVER_ID": "auto"
      }
    }
  }
}

Prompts to try

Use these prompts when exploring the demos or integrating CapiscIO security into your own MCP servers.

- "Run the enforcement demo and show me which agents are allowed vs denied for the place_order tool"
- "Explain the trust badge verification result for the badged agent"
- "Switch the active policy to Lockdown and show me what happens to tool access"
- "Show the DID and badge for the LangChain research agent"

Troubleshooting A2A Demos

setup.sh fails to download the required binary

Run ./setup.sh before going offline — the script pre-downloads a ~15 MB binary required by all demos. If you are behind a proxy or firewall, ensure outbound HTTPS to capisc.io is permitted.

CAPISCIO_API_KEY not found or authentication errors

Get your API key from app.capisc.io → Settings → API Keys. Copy it exactly into the .env file in the relevant demo directory. For CAPISCIO_SERVER_ID, set it to 'auto' to have CapiscIO create one automatically on first run.

Multi-agent demo agents fail to start or cannot reach each other

Each agent needs its own terminal with the correct virtualenv activated. Ports 8001, 8002, and 8003 must be free. Run `python run_demo.py --discover` to check which agents are reachable before sending tasks.

Frequently Asked Questions about A2A Demos

What is A2A Demos?

A2A Demos is a Model Context Protocol (MCP) server that demo agents showcasing capiscio agent guard and mcp guard — trust badges, identity verification, and tool-level authorization for a2a and mcp protocols It connects AI assistants to external tools and data sources through a standardized interface.

How do I install A2A Demos?

Follow the installation instructions on the A2A Demos GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with A2A Demos?

A2A Demos works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is A2A Demos free to use?

Yes, A2A Demos is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

A2A Demos Alternatives — Similar Security Servers

Looking for alternatives to A2A Demos? Here are other popular security servers you can use with Claude, Cursor, and VS Code.

Casdoor

13.6k

An open-source Agent-first Identity and Access Management (IAM) /LLM MCP & agent gateway and auth server with web UI supporting OpenClaw, MCP, OAuth, OIDC, SAML, CAS, LDAP, SCIM, WebAuthn, TOTP, MFA, Face ID, Google Workspace, Azure AD

ghidraMCP

9.0k

An Model Context Protocol server that enables LLMs to autonomously reverse engineer applications by exposing Ghidra's decompilation and analysis tools. It allows AI agents to list code structures, rename methods, and analyze binaries directly through

HexStrike AI

8.9k

HexStrike AI MCP Agents is an advanced MCP server that lets AI agents (Claude, GPT, Copilot, etc.) autonomously run 150+ cybersecurity tools for automated pentesting, vulnerability discovery, bug bounty automation, and security research. Seamlessly b

IDA Pro MCP

8.7k

Enables AI-assisted reverse engineering in IDA Pro by providing tools to analyze binaries, decompile functions, manage comments, search patterns, and interact with the IDA database through natural language.

Anthropic Cybersecurity Skills

6.6k

754 structured cybersecurity skills for AI agents · Mapped to 5 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND & NIST AI RMF · agentskills.io standard · Works with Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI & 20+ platform

Hooker

5.1k

🔥🔥 hooker is a Frida-based reverse engineering toolkit for Android. It offers a user-friendly CLI, universal scripts, auto hook generation, memory roaming to detect activities/services, one-click SOCKS5 proxy setup, Frida JustTrustMe, and BoringSSL u

Browse More Security MCP Servers

Explore all security servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "a2a-demos": { "command": "npx", "args": ["-y", "a2a-demos"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use A2A Demos?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides