MCP

Connect your AI assistant to on-chain market tools.

The Mattheus MCP surface exposes account-aware research, positions, balances, markets, backtests, algorithms, and guarded workflow tools according to the certified action registry. Public execution tools remain disabled while the global verdict is NO_GO.

Account context

Give compatible assistants structured access to positions, balances, markets, protocols, and algorithm status.

Tooling for agents

Use MCP as the bridge between AI assistants and Mattheus research, simulation, and workflow capabilities.

Governed execution

Execution tooling remains scoped, approval-aware, and clear about risk gates before any enabled live action.

Research-only

Use the assistant to analyze markets, explain protocols, create artifacts, and prepare workflows without enabling live execution.

Approval-first

Generate execution previews and approval requests. Live actions stay blocked until user approval and policy gates pass.

Approval-first execution

Run deployed algorithms only after configured risk limits, signer checks, venue gates, stale-data stops, and approval policies pass.

Related Mattheus paths

Common Questions

What does MCP expose?

MCP can expose tools for algorithms, positions, portfolio data, balances, rates, markets, protocols, backtests, and account operations depending on configured access.

Can an external assistant place trades by itself?

Execution should stay governed by API scopes, reference IDs, signer readiness, subscriptions, venue gates, and approval policy.

Can developers use Mattheus without MCP?

Yes. Mattheus also exposes REST API, Python SDK, and TypeScript SDK surfaces for read workflows, previews, and governed execution where enabled.

Where are the detailed MCP setup steps?

The detailed setup remains in the Mattheus MCP developer docs.

Mattheus provides research, simulation, automation, and execution tooling for on-chain markets. It does not promise outcomes or provide individualized investment advice.