Pit Officer
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Pit OfficerValidate order 12345 against pre-trade risk limits and sign if approved."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Pit Officer
Pit Officer is a control plane for the embeddable OpenPit pre-trade risk engine. It runs one engine in a local process over a local SQLite database and exposes it to operators through a web dashboard, to programs through a REST API, and to AI agents through an MCP server. The engine alone evaluates orders.
Capabilities
One local engine over one local store - a single node built from SQLite at startup.
A reusable framework: the importable Go packages -
go.openpit.dev/officerwith itsweb,httpapi,mcptools,engine,signing, andframework/...packages - andweb/src/framework/in React. This application is one composition of them; a program can build its own.A REST API over the control plane - see the API notes.
An MCP server over stdio or streamable HTTP, with mutating tools off until an operator enables them by name.
An operator dashboard, an embedded single-page app driven by the same API.
Signed order approvals - an Ed25519 approval envelope for every order that passes pre-trade.
An append-only audit trail of every control-plane action.
Market data through pluggable connectors, configured and toggled at runtime.
Optional at-rest encryption of stored secrets under an operator-supplied master key.
Related MCP server: AgentGuard MCP Server
Run
Pit Officer is a single binary with four subcommands: serve, mcp,
dashboard, healthcheck. With Just:
just run-serve # build the SPA and the binary, then serve
just dashboard # print the running instance's URLs and open the browser
just run-mcp # local stdio MCP serverIn a container, build the image with just docker-build and start it with
docker compose up; docker-compose.yml pins the port, the database volume,
and a read-only root filesystem.
The subcommands, the environment variables and flags, and the authorization model are in the configuration reference.
Build
Prerequisites: a C toolchain with cgo enabled, which the OpenPit Go binding
requires; Go; Node and npm, for the dashboard bundle; Python 3, which the
recipes use to drive Go, the linters, and Semgrep; golangci-lint; and
Just. Tool versions are pinned in
.github/ci-versions.env.
just frontend-ci-install # npm ci in web/ from the committed lockfile
just build # build the SPA, then the pit-officer binary
just check # format checks, linters, Semgrep, tests, frontend bundlesComposing your own Officer from the importable packages, building against a local Pit checkout, building without Just, and the debug variants are covered in the development notes.
License
Apache License 2.0. See LICENSE.
This server cannot be deployed
Maintenance
Related MCP Connectors
Zero-secret MCP gateway for AI agents: risk-scored, audited calls with human-in-the-loop approval.
Read-only checks for proposed orders against evidence and policy; no trading or payment authority.
Deterministic pre-trade risk checks for autonomous AI trading agents.
Read-only finance and operations controls for AI agents with evidence and safe next actions.
Related MCP Servers
AlicenseNot gradedqualityAmaintenanceEnables AI agents to securely call MCP tools with risk scoring, checkpoints, rollback, and approval workflows.1 npmMIT- FlicenseNot gradedqualityBmaintenanceProvides a secure MCP boundary for AI agents, intercepting and validating tool calls, redacting secrets, and requiring human approval for sensitive actions with a tamper-evident audit trail.-
- FlicenseNot gradedqualityCmaintenanceMCP server that provides a security gateway for AI agents, enforcing allow/confirm/deny policies on tool calls and requiring human approval for risky operations, with full audit logging.-
- AlicenseNot gradedqualityBmaintenanceEnables MCP-compatible AI agents to safely act on business backends by enforcing per-agent permissions, autonomy thresholds, human approval with review-and-edit, and full audit trails.MIT