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
Website: officer.openpit.dev
Pit Officer lets you manage trading accounts, set risk limits, connect to brokers, and send orders from a web panel. The OpenPit risk engine checks each order before Officer sends it through a trading connection. You can also use Officer from trading software or AI agents.
Run
For a visual walkthrough, see the video guides.
Officer runs on your own computer, and you use it through a web browser. Docker lets you run the same ready-made application on Windows, macOS, and Linux, without building it or installing programming tools.
If you do not have Docker yet, follow the Docker installation guide. Start Docker Desktop and wait until it says it is running. On Linux, you can also use Docker Engine; see the quickstart.
Open a terminal, the window where you type commands: PowerShell from the Start menu on Windows, Terminal on macOS, or your terminal application on Linux. Copy the command below, paste it there, and press Enter. The command is the same on all three systems:
docker run -d --name pit-officer --restart unless-stopped -p 127.0.0.1:8787:8787 -v pit-officer-data:/data ghcr.io/openpitkit/officer:latestlatest is the newest stable release of Officer. The first run downloads
Officer, which may take a few minutes. It then runs in the background, so you
can close the terminal. Your settings and records are saved on your computer
in storage named pit-officer-data and kept when you stop or restart Officer.
To check whether it is ready, run:
docker ps --filter name=pit-officerWhen the STATUS column shows (healthy), open http://127.0.0.1:8787 in your
browser and bookmark it. If it shows (health: starting), wait a few seconds
and check again. The panel is ready for you to configure your accounts, risk
limits, and connections.
In that address, 127.0.0.1 means your own computer, and 8787 is the port,
the number used to reach the panel. To use a different port, choose it before
running the long command: for example, replace -p 127.0.0.1:8787:8787 with
-p 127.0.0.1:8788:8787, then open http://127.0.0.1:8788. Keep the final
8787 unchanged; it is the number Officer uses inside Docker.
To stop Officer, run docker stop pit-officer. To start it again, run
docker start pit-officer and open your bookmark. Use these commands after
the first installation; you do not need to repeat the long command.
To try changes before they are released, run the edge build instead: the
newest version under development that passed its checks. It changes often and
may not work, so give it its own name, port, and storage, apart from your
stable installation:
docker run -d --name pit-officer-edge --restart unless-stopped -p 127.0.0.1:8788:8787 -v pit-officer-edge-data:/data ghcr.io/openpitkit/officer:edgeThen open http://127.0.0.1:8788, and use pit-officer-edge instead of
pit-officer in the commands above.
For help with installation, changing the port of an existing installation, or a panel that will not open, see the quickstart.
Related MCP server: agent-security-gateway
Build
Prerequisites: a C toolchain, which the OpenPit Go binding requires through
cgo; Go; and Node and npm for the dashboard bundle. Tool versions are pinned in
.github/ci-versions.env.
From the repository root, build the dashboard before the binary so Go embeds the compiled assets:
npm --prefix web ci
npm --prefix web run build
CGO_ENABLED=1 go build -o pit-officer ./cmd/pit-officerOn Windows, set CGO_ENABLED=1 in your shell, build with
go build -o pit-officer.exe ./cmd/pit-officer, and run ./pit-officer.exe.
Run from source
Pit Officer is a single binary with four subcommands: serve, mcp,
dashboard, healthcheck. After building it, run from the repository root:
./pit-officer serve # serve the dashboard, REST API, and HTTP MCP
./pit-officer dashboard # in another terminal, open the running instance
./pit-officer mcp # alternatively, run a local stdio MCP serverThe subcommands, environment variables, flags, and authorization model are in the configuration reference. See the API notes for REST and MCP integration.
Local container build
To build and run a local container image with the settings in
docker-compose.yml:
. ./.github/ci-versions.env
docker build --build-arg GO_VERSION="$CI_GO" -t pit-officer:local .
docker compose upOptional Just shortcuts
Just provides a quicker alternative to the manual steps above. The recipes also require Python 3; the full check additionally uses golangci-lint and Semgrep, as described in the development notes.
just frontend-ci-install # npm ci in web/ from the committed lockfile
just build # build the SPA, then the pit-officer binary
just run-serve # build and serve
just dashboard # open the running instance in the browser
just run-mcp # build and run a local stdio MCP server
just docker-build # build the pit-officer:local container image
just check # format checks, linters, Semgrep, tests, frontend bundlesComposing your own Officer from the importable packages, building against a local Pit checkout, 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.
Free, read-only directory of 114 AI trading tools: broker MCP servers, market data APIs, bots and backtesters. Sourced facts on who can place orders and what safeguards exist; search, compare and check order gating for every broker MCP.
Deterministic pre-trade risk checks for AI agents on Robinhood Chain. Returns allow/warn/deny.
Related MCP Servers
AlicenseNot gradedqualityBmaintenanceEnables AI agents to securely call MCP tools with risk scoring, checkpoints, rollback, and approval workflows.17 npmMIT- FlicenseNot gradedqualityDmaintenanceMCP 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.-
- AlicenseAqualityBmaintenanceEnables AI agents to place and manage Hyperliquid trades through MCP with position, leverage, daily-loss, market, and stop-loss limits enforced outside the model, plus paper trading and an auditable decision log.955 npmMIT
- FlicenseNot gradedqualityBmaintenanceEnables autonomous agents and MCP-compatible clients to enforce policy-approved actions using risk tiers, reversible checkpoints, and token approval bounds.7-