EVD2.ca is Now Agent-Ready: MCP + A2A Support

EVD2.ca now supports MCP and A2A, the two protocols AI agents are converging on for data access and inter-agent collaboration. MCP turns EVD2 into a native data source your AI can query directly for Canadian EV specs, news, and rebate information. A2A lets other agents discover EVD2 as a specialist and delegate EV questions to it. The result: your AI assistant can research Canadian EVs, track updates, and check incentive eligibility without you ever opening a browser tab.

The way people find information is changing fast.

Search isn't just Google anymore. It's ChatGPT, Claude, Perplexity, and increasingly, AI agents acting on your behalf.

So we've been building toward that future.

Today, EVD2.ca officially supports:

• MCP (Model Context Protocol)
• A2A (Agent-to-Agent communication)

And it changes how you interact with EV data.

From "Search" to "Ask and Act"

If you wanted to research an EV the old way, you'd open a browser, visit a handful of sites, compare specs by hand, and track changes yourself.

Now you can stay inside your AI assistant and just ask:

"Track new EVs eligible for Canadian rebates." "Notify me when a new Chinese EV is approved for sale in Canada." "Compare all AWD EV SUVs under $60K."

And your AI can pull structured data directly from EVD2, track updates over time, and act on your behalf.

No tabs. No context switching.

What MCP Enables (Agent to Data)

At the core of this is MCP (Model Context Protocol).

MCP is an open standard that lets AI models connect directly to tools, APIs, and structured data sources without custom integrations for each one.

In simple terms: it turns EVD2 into a native data source for AI.

Instead of scraping pages or guessing intent, agents can query EV data cleanly, access structured vehicle specs, and understand updates as they happen.

MCP solves the last-mile problem of AI: giving it real, trusted context from live systems instead of stale training data.

What A2A Enables (Agent to Agent)

Then comes the next layer: A2A (Agent-to-Agent).

If MCP is about connecting AI to tools, A2A is about connecting AI to other AI.

A2A is an open protocol designed for agents to discover each other, share capabilities, collaborate on tasks, and delegate work across systems.

The distinction is simple:

• MCP = "use this data/tool"
• A2A = "work with this other agent"

Or even simpler: MCP is capability access. A2A is collaboration.

Together, they unlock multi-agent workflows where your personal AI doesn't just fetch data. It coordinates outcomes.

What This Means for EVD2

EVD2 is no longer just a website. It's becoming an agent-accessible EV intelligence layer.

That means:

AI can track vehicles for you. Instead of manually checking updates, your agent can follow specific EV models, monitor regulatory approvals, and watch pricing and availability changes.

You can build your own EV agent. Using MCP and A2A, you can create a personal EV assistant, connect it to EVD2, and combine it with other tools like pricing engines, dealership inventory, and incentive databases.

Zero UI required. Everything can happen directly inside ChatGPT, Claude, or any agent framework. You don't need to visit EVD2. Your AI brings it to you.

Why This Matters

This shift is as big as mobile. Maybe bigger.

We're moving from websites to APIs, APIs to AI tools, and AI tools to autonomous agents.

And here's the part most people aren't ready for: the best products of the next decade won't be the ones with the best UI. They'll be the ones AI can use best. Design for humans is becoming a secondary concern. Design for agents is the new game.

Protocols like MCP and A2A are becoming the infrastructure layer of the AI-native internet.

100/100 Agent Readiness

We've been actively optimizing EVD2 for this future.

Full audit here: isitagentready.com/www.evd2.ca

The score reflects structured, machine-readable content, clean agent access patterns, and compatibility with the protocols agents are converging on.

What's Next

This is just the start. We're exploring agent-based EV alerts (push, not pull), personalized EV recommendation agents, multi-agent workflows that combine finance, vehicle, and incentive data, and direct agent integrations with dealerships and inventory systems.

Try It

If you're building with agents, connect your agent to EVD2 via MCP, experiment with A2A workflows, and build something useful.

If you're just curious, ask your AI something about EVs in Canada and see how far it gets.

Final Thought

We're entering a world where you don't browse for information. Your agent negotiates reality on your behalf.

EVD2 is being built for that world.

Frequently Asked Questions

What is MCP? MCP (Model Context Protocol) is an open standard that lets AI models connect directly to structured data sources. Instead of scraping websites or relying on training data, AI agents can query EVD2's vehicle database, news feed, and rebate information through a clean, typed interface. It gives AI native access to your data instead of making it guess.

What is A2A? A2A (Agent-to-Agent) is an open protocol for AI agents to discover each other, share capabilities, and delegate tasks. When your AI needs Canadian EV data, it can find EVD2 through a standard discovery URL and send structured queries. MCP is about data access. A2A is about agent collaboration.

Do I need to do anything differently to use this? No. If you use ChatGPT, Claude, Perplexity, or any AI assistant that supports MCP or A2A, it can already discover and query EVD2. Ask your AI about Canadian EVs and it handles the rest. No setup, no API keys, no configuration on your end.

What data can AI agents access through EVD2? Three things: vehicle specs (every EV approved for sale in Canada with pricing, range, battery, and availability), Canadian EV news (articles from 30+ sources matched to specific vehicles), and rebate information (federal EVAP and provincial incentives by province). All read-only. No personal data, no admin access, no write operations.

Is my data safe? The A2A and MCP endpoints are strictly read-only. They expose the same public vehicle, news, and rebate data already visible on the website. No user accounts, no personal information, and no admin functionality is accessible through these protocols. Rate limiting prevents abuse.

What does a 100/100 agent readiness score mean? EVD2 scores a perfect 100 on isitagentready.com, which audits websites for machine readability, structured data, agent access patterns, and protocol compatibility. The score reflects clean structured content, proper discovery endpoints, and compatibility with the protocols AI agents are converging on.

Can I build my own EV agent using EVD2? Yes. EVD2's MCP and A2A endpoints are open. Connect any MCP-compatible tool or A2A-capable agent to EVD2 and build custom workflows: look up vehicles, monitor news, check rebates, and combine it with other data sources like pricing engines or dealership inventory.

What is the difference between using EVD2 through an AI agent versus visiting the website? The data is the same. The difference is how you access it. On the website, you browse and search manually. Through an agent, your AI pulls exactly what it needs, combines it with other sources, and gives you a synthesized answer. No tabs, no context switching. The agent does the legwork.

Colin