AI Agent Integration

Agent-Native Cyber Insurance for Canadian Micro-Businesses

Connect your AI agents directly to CyberAgency via the Model Context Protocol (MCP). Assess cyber risk, generate quotes, explain coverage, and initiate policy binding — all from within Claude, ChatGPT, Zapier AI, or any MCP-compatible agent platform.

Free Risk Assessment
No Signup for Open Tools
Human-Approval Bind Flow
MCP Server — Live
AI Agent (Claude / GPT)
MCP Protocol (SSE / stdio)
assess_risk → quote → bind
No auth needed
5 MCP tools

How Agent Integration Works

Your AI agent calls CyberAgency tools through the Model Context Protocol — no custom API integration required.

AI Agent
Claude, ChatGPT, Zapier
MCP Protocol
HTTP/SSE or stdio
CyberAgency Tools
Risk, Quote, Coverage, Bind
1

Assess Risk

Agent calls assess_risk with business details — industry, size, tools, AI usage, and data flows.

2

Generate Quote

Agent calls quote_cyber with the risk profile to get annual/monthly premium and coverage limit.

3

Explain Coverage

Agent calls explain_coverage for a plain-language summary of what's covered and what's excluded.

4

Initiate Bind

Agent calls initiate_bind (optional). Human approves via secure URL before anything is finalized.

MCP Tools Reference

Five tools your agents can call. Three are open and free. Two require an API key for transaction accountability.

assess_risk Open

Assess cyber risk for a Canadian business based on operational details. Returns a risk score, identified gaps, and recommendations.

Inputs industry, employee_count, tools, ai_usage, data_flows
quote_cyber Open

Generate a cyber insurance quote from a risk profile. Returns annual premium, monthly premium, and coverage limit.

Inputs risk_profile (from assess_risk), coverage_tier (optional)
explain_coverage Open

Get a plain-language explanation of coverage for a given quote. Helps agents communicate value to business owners.

Inputs quote_id (from quote_cyber)
initiate_bind API Key

Initiate a policy binding request. Generates a secure approval URL where a human reviews and confirms before binding.

Inputs quote_id, business_details, contact_email
check_bind_status API Key

Check the current status of a bind request. Returns whether it's pending, approved, declined, or completed.

Inputs bind_id (from initiate_bind)

Quick Start

Connect your agent in under five minutes. Choose production HTTP/SSE or local stdio for development.

HTTP/SSE

Production
// Claude Desktop config (claude_desktop_config.json)
{
  "mcpServers": {
    "cyberagency": {
      "url": "https://mcp.cyberagency.ca/sse"
    }
  }
}

stdio

Local Dev
# Clone and run locally via stdio
git clone https://github.com/barnaby-coder/cyberagency-mcp
cd cyberagency-mcp && npm install
npm run dev

Integration Examples

Works with any MCP-compatible platform. Here are three popular options.

Claude Desktop

Add CyberAgency to your Claude Desktop config and start assessing risk immediately. No code required.

// claude_desktop_config.json
{
  "mcpServers": {
    "cyberagency": {
      "url": "https://mcp.cyberagency.ca/sse"
    }
  }
}

Zapier AI

Use CyberAgency MCP tools inside Zapier AI workflows. Automate risk checks and quote generation for leads.

// Zapier AI Actions config
{
  "mcp_endpoint": "https://mcp.cyberagency.ca/sse",
  "tools": ["assess_risk", "quote_cyber"]
}

Python MCP Client

Build a custom integration using the MCP Python SDK. Full control over tool selection and response handling.

# pip install mcp
from mcp import ClientSession

async with ClientSession(
    "https://mcp.cyberagency.ca/sse"
) as session:
    result = await session.call_tool(
        "assess_risk",
        {"industry": "legal", "employee_count": 15}
    )

Authentication Model

Open tools for discovery. Authenticated tools for transactions. Simple and predictable.

● Open — No Auth Required

Top-of-Funnel Tools

These tools are free to use with no signup. They're designed for discovery — let your agents explore cyber risk for any Canadian business.

  • assess_risk — generate a risk profile
  • quote_cyber — get pricing estimates
  • explain_coverage — plain-language summaries
● Bearer Token Required

Transaction Tools

These tools involve binding and policy actions. A bearer token ensures accountability and creates an audit trail for each transaction.

  • initiate_bind — start the binding process
  • check_bind_status — track bind progress
Get an API Key →

MCP Integration FAQ

Short, citation-friendly answers for founders, brokers, and agent builders evaluating CyberAgency integration.

What is the CyberAgency MCP server?

It is an MCP endpoint that gives AI agents direct access to CyberAgency cyber insurance workflows. Agents can assess cyber risk, generate quotes, explain coverage in plain language, and hand off a bind request for human approval.

Which tools are open versus authenticated?

assess_risk, quote_cyber, and explain_coverage are open for top-of-funnel discovery. initiate_bind and check_bind_status require a bearer token because they involve transaction accountability and approval tracking.

Who is this built for?

CyberAgency MCP is designed for Canadian micro-businesses, brokers, AI workflow builders, and platforms that want cyber insurance recommendations or quoting embedded directly inside agent experiences.

How do I connect my agent?

Use the production SSE endpoint at https://mcp.cyberagency.ca/sse from an MCP-compatible client such as Claude Desktop, Zapier AI, or a custom SDK integration. Pilot access for bind tools is available at [email protected].

Join the MCP Pilot Program

Be among the first to connect AI agents to Canadian cyber insurance. Limited spots available for our pilot program.