QuantConnect
OfficialConnect an AI assistant to QuantConnect to manage projects, code, backtests, optimizations, live deployments, and related resources.
Read organization/account status.
Create, list, read, update, delete, and manage projects and collaborators; lock projects for editing.
Read/update project nodes and compile projects.
Create, read, rename, patch, update contents of, and delete project files.
Create, read, list, update, and delete backtests; inspect charts, orders, and insights.
Estimate, create, read, list, update, abort, and delete optimizations.
Authorize brokerage connections and create/read/list/stop/liquidate live algorithms.
Read live charts, logs, portfolio, orders, and insights; send or broadcast live commands.
Upload, list, inspect, download, and delete Object Store files.
Read LEAN versions and use code assistance tools for completion, syntax checks, PEP8 formatting, error enhancement, initialization checks, and QuantConnect search.
Check MCP server version information.
Provides comprehensive access to QuantConnect's algorithmic trading platform, enabling AI agents to create and manage trading projects, write and backtest strategies, run optimizations, deploy algorithms to live trading, manage files and collaborators, and interact with the Object Store for data management.
Click 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., "@QuantConnect create a new project for my momentum trading strategy"
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.
QuantConnect Remote MCP Server
The QuantConnect Remote MCP Server is the preferred way to connect AI assistants such as Claude to the QuantConnect platform.
There is nothing to install, build, or run. Point your MCP client at the hosted endpoint, authorize it against your QuantConnect organization, and the tools are available immediately.
Documentation: https://www.quantconnect.com/docs/v2/ai-assistance/mcp-server/key-concepts
Remote MCP endpoint:
https://www.quantconnect.com/api/v2/mcp
Note: The remote MCP server requires a paid agent node (A1-1, A4-8, or A16-32) in your organization. The free A-MICRO node cannot connect. See Agent Nodes and Add Nodes.
Quick Start with Claude
The remote server speaks the standard MCP HTTP transport, so any MCP-capable Claude surface can use it. Pick the one you use.
Claude Desktop and claude.ai
Open Settings > Connectors (on the desktop app, File > Settings > Connectors).
Click Add custom connector, give it a name such as
QuantConnect, and enter the server URL:https://www.quantconnect.com/api/v2/mcpClick Connect. Your default browser opens.
Click Continue connecting, select your QuantConnect organization in the Authorize MCP Server form, and click Authorize.
Start a new chat. The QuantConnect tools are listed under the connector and can be enabled per conversation.
Claude Code (terminal)
Register the server once from any terminal:
claude mcp add --transport http quantconnect https://www.quantconnect.com/api/v2/mcpThen start Claude Code, run /mcp, select quantconnect, and follow the browser authorization flow (Continue connecting > Authorize). Running /mcp again should show the server as connected with its tools listed.
Claude Code in Local Platform (VS Code extension)
Install the Claude Code extension in VS Code.
Create or open a QuantConnect project in Local Platform.
Open the Claude Code panel from the top navigation bar.
Enter
/mcpto confirm the QuantConnect MCP Server is connected.
Try it
Once connected, send a simple prompt to verify the tools are working:
Read the open project.Claude should call the read_open_project tool and describe the project. From there you can ask it to create projects, write and edit strategy files, run backtests, optimize parameters, or deploy to paper trading, all in plain language. For example:
Create a new Python project that buys SPY on the first trading day of each month,
backtest it from 2020 to today, and summarize the results.Tip for Claude Code users: a CLAUDE.md file in your project or home directory lets you set conventions (Python-first, compile after every edit, position-sizing rules, and so on). The Claude Code docs page includes an example you can copy.
Related MCP server: CCXT MCP Server
Available Tools
The full, up-to-date list lives in the Key Concepts documentation.
Other Clients
The remote server also works with Copilot, Cursor, Devin Desktop, and any other client that supports the MCP HTTP transport. Setup guides for each are in the MCP Server documentation.
Available Tools
64 toolsabort_optimizationCIdempotent
Abort an optimization.
Inputs: model (required). Returns: success, errors
authorize_connectionAIdempotent
Authorize an external connection with a live brokerage or data provider.
This tool automatically opens your browser for you to complete the authentication flow. For the flow to work, you must be logged into your QuantConnect account on the browser that opens.
Inputs: model (required). Returns: authorization, success, errors
broadcast_live_commandC
Broadcast a live command to all live algorithms in an organization.
Inputs: model (required). Returns: success, errors
check_initialization_errorsBRead-only
Run a backtest for a few seconds to initialize the algorithm and get inialization errors if any.
Inputs: model (required). Returns: state, version, payload, payloadType
check_syntaxCRead-only
Check the syntax of a code.
Inputs: model (required). Returns: state, version, payload, payloadType
complete_codeCRead-only
Show the code completion for a specific text input.
Inputs: model (required). Returns: state, version, payload, payloadType
create_backtestC
Create a new backtest request and get the backtest Id.
Inputs: model (required). Returns: backtest, debugging, success, errors
create_compileB
Asynchronously create a compile job request for a project.
Inputs: model (required). Returns: compileId, state, parameters, projectId, +4 more
create_fileAIdempotent
Add a file to a given project.
Inputs: model (required). Returns: success, errors
create_live_algorithmD
Create a live algorithm.
Inputs: model (required). Returns: responseCode, source, deployId, versionId, +4 more
create_live_commandC
Send a command to a live trading algorithm.
Inputs: model (required). Returns: success, errors
create_optimizationC
Create an optimization with the specified parameters.
Inputs: model (required). Returns: optimizations, success, errors
create_projectB
Create a new project in your default organization.
Inputs: model (required). Returns: projects, versions, success, errors
create_project_collaboratorBIdempotent
Add a collaborator to a project.
Inputs: model (required). Returns: collaborators, success
delete_backtestAIdempotent
Delete a backtest from a project.
Inputs: model (required). Returns: success, errors
delete_fileBIdempotent
Delete a file in a project.
Inputs: model (required). Returns: success, errors
delete_objectBIdempotent
Delete the Object Store file of a specific organization and key.
Inputs: model (required). Returns: success, errors
delete_optimizationBIdempotent
Delete an optimization.
Inputs: model (required). Returns: success, errors
delete_projectCIdempotent
Delete a project.
Inputs: model (required). Returns: success, errors
delete_project_collaboratorCIdempotent
Remove a collaborator from a project.
Inputs: model (required). Returns: collaborators, success
enhance_error_messageBRead-only
Show additional context and suggestions for error messages.
Inputs: model (required). Returns: state, version, payload, payloadType
estimate_optimization_timeCRead-only
Estimate the execution time of an optimization with the specified parameters.
Inputs: model (required). Returns: estimate, success, errors
liquidate_live_algorithmCIdempotent
Liquidate and stop a live algorithm.
Inputs: model (required). Returns: success, errors
list_backtestsARead-only
List all the backtests for the project.
Inputs: model (required). Returns: backtests, count, success, errors
list_live_algorithmsC
List all your past and current live trading deployments.
Inputs: model (required). Returns: live, success, errors
list_object_store_filesBRead-only
List the Object Store files under a specific directory in an organization.
Inputs: model (required). Returns: path, objects, page, totalPages, +4 more
list_optimizationsARead-only
List all the optimizations for a project.
Inputs: model (required). Returns: optimizations, success, errors
list_projectsARead-only
List the details of all projects.
Inputs: no named parameters. Returns: projects, versions, success, errors
lock_project_with_collaboratorsAIdempotent
Lock a project so you can edit it.
This is necessary when the project has collaborators or when an LLM is editing files on your behalf via our MCP Server.
Inputs: model (required). Returns: success, errors
patch_fileAIdempotent
Apply a patch (unified diff) to a file in a project.
Inputs: model (required). Returns: success, errors
read_accountARead-only
Read the organization account status.
Inputs: no named parameters. Returns: organizationId, creditBalance, card
read_backtestBRead-only
Read the results of a backtest.
Inputs: model (required). Returns: backtest, debugging, success, errors
read_backtest_chartCRead-only
Read a chart from a backtest.
Inputs: model (required). Returns: chart, success, errors
read_backtest_insightsCRead-only
Read out the insights of a backtest.
Inputs: model (required). Returns: insights, length, success
read_backtest_ordersBRead-only
Read out the orders of a backtest.
Inputs: model (required). Returns: orders, length
read_compileBRead-only
Read a compile packet job result.
Inputs: model (required). Returns: compileId, state, logs, success, +1 more
read_fileARead-only
Read a file from a project, or all files in the project if no file name is provided.
Inputs: model (required). Returns: files, success, errors
read_latest_mcp_server_versionBRead-only
Returns the latest version of the QC MCP Server released.
Inputs: no named parameters. Returns: result
read_lean_versionsARead-only
Returns a list of LEAN versions with basic information for each version.
Inputs: no named parameters. Returns: versions, success, errors
read_live_algorithmC
Read details of a live algorithm.
Inputs: model (required). Returns: message, deployId, status, cloneId, +13 more
read_live_chartC
Read a chart from a live algorithm.
Inputs: model (required). Returns: chart, success, errors
read_live_insightsB
Read out the insights of a live algorithm.
The snapshot updates about every 10 minutes.
Inputs: model (required). Returns: insights, length, success
read_live_logsA
Get the logs of a live algorithm.
The snapshot updates about every 5 minutes.
Inputs: model (required). Returns: logs, length, deploymentOffset, success, +1 more
read_live_ordersA
Read out the orders of a live algorithm.
The snapshot updates about every 10 minutes.
Inputs: model (required). Returns: orders, length, success, errors
read_live_portfolioB
Read out the portfolio state of a live algorithm.
The snapshot updates about every 10 minutes.
Inputs: model (required). Returns: portfolio, success, errors
read_mcp_server_versionARead-only
Returns the version of the QC MCP Server that's running.
Inputs: no named parameters. Returns: result
read_object_propertiesARead-only
Get Object Store properties of a specific organization and key.
It doesn't work if the key is a directory in the Object Store.
Inputs: model (required). Returns: metadata, success, errors
read_object_store_file_download_urlBRead-only
Get the URL for downloading files from the Object Store.
Inputs: model (required). Returns: jobId, url, success, errors
read_object_store_file_job_idC
Create a job to download files from the Object Store and then read the job Id.
Inputs: model (required). Returns: jobId, url, success, errors
read_optimizationCRead-only
Read an optimization.
Inputs: model (required). Returns: optimization, success, errors
read_projectBRead-only
List the details of a project or a set of recent projects.
Inputs: model (required). Returns: projects, versions, success, errors
read_project_collaboratorsARead-only
List all collaborators on a project.
Inputs: model (required). Returns: collaborators, userLiveControl, userPermissions, success
read_project_nodesARead-only
Read the available and selected nodes of a project.
Inputs: model (required). Returns: nodes, autoSelectNode, success, errors
search_quantconnectBRead-only
Search for content in QuantConnect.
Inputs: model (required). Returns: state, version, retrivals, messageId
stop_live_algorithmBIdempotent
Stop a live algorithm.
Inputs: model (required). Returns: success, errors
update_backtestAIdempotent
Update the name or note of a backtest.
Inputs: model (required). Returns: success, errors
update_code_to_pep8CRead-only
Update Python code to follow PEP8 style.
Inputs: model (required). Returns: state, version, payload, payloadType
update_file_contentsCIdempotent
Update the contents of a file.
Inputs: model (required). Returns: files, success, errors
update_file_nameBIdempotent
Update the name of a file.
Inputs: model (required). Returns: success, errors
update_optimizationBIdempotent
Update the name of an optimization.
Inputs: model (required). Returns: success, errors
update_projectAIdempotent
Update a project's name or description.
Inputs: model (required). Returns: success, errors
update_project_collaboratorBIdempotent
Update collaborator information in a project.
Inputs: model (required). Returns: collaborators, success
update_project_nodesCIdempotent
Update the active state of the given nodes to true.
If you don't provide any nodes, all the nodes become inactive
and autoSelectNode is true.
Inputs: model (required). Returns: nodes, autoSelectNode, success, errors
upload_objectCIdempotent
Upload files to the Object Store.
Inputs: model (required). Returns: success, errors
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 64 tools
Tools are mostly distinct, with clear read/create/update/delete boundaries per resource domain. Minor overlap exists between list_projects and read_project (both can list projects), and stop_live_algorithm vs liquidate_live_algorithm could be misselected by an agent not reading carefully.
Most tools follow a consistent verb_noun snake_case pattern (create_backtest, read_backtest, list_backtests, delete_file). Deviations include object vs object_store prefixes, read_object_store_file_job_id which actually creates a job, and read_lean_versions used for listing.
This server exposes 64 tools, well above the 25+ threshold the rubric flags as too many for typical agent use. Even though the broad platform has genuine sub-domains, the sheer number makes the set heavy to navigate and increases the chance of misselection.
The tool set covers most core lifecycle workflows: projects, files, compiles, backtests, live algorithms, optimizations, object store, and collaboration. A few minor gaps exist, such as no ability to update optimization parameters after creation, no backtest abort, and no live algorithm update, but these are workable.
Maintenance
Related MCP Connectors
Official Intrinio MCP server. US stock prices, options and fundamentals for AI agents.
MCP server for Financial Modeling Prep API with 250+ financial data tools
The official MCP Server for the Mux API
Related MCP Servers
- MIT
- AlicenseBqualityDmaintenanceHigh-performance CCXT MCP server for cryptocurrency exchange integration2494 npm145MIT
- GoMIT
- MIT