QuantConnectSAFE
Official Python MCP server for local interactions with the QuantConnect API
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
See https://www.quantconnect.com/docs/v2/ai-assistance/mcp-server/key-concepts
QuantConnect MCP Server
The QuantConnect MCP Server is a bridge for AIs (such as Claude and OpenAI o3 Pro) to interact with our cloud platform. When equipped with our MCP, the AI can perform tasks on your behalf through our API such as updating projects, writing strategies, backtesting, and deploying strategies to production live-trading.
This is the OFFICIAL implementation of QuantConnect's MCP, maintained by the QuantConnect team. We recommend using the official version to ensure security of your code and API tokens. Our implementation is tested and dockerized for easy cross-platform deployment.
Getting Started
To connect local MCP clients (like Claude Desktop) to the QC MCP Server, follow these steps:
- Install and open Docker Desktop.
- Install and open Claude Desktop.
- In Claude Desktop, click File > Settings > Developer > Edit Config.
- Edit the
claude_desktop_config.jsonfile to include the followingquantconnectconfiguration:
{
"mcpServers": {
"quantconnect": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "QUANTCONNECT_USER_ID",
"-e", "QUANTCONNECT_API_TOKEN",
"-e", "AGENT_NAME",
"--platform", "",
"quantconnect/mcp-server"
],
"env": {
"QUANTCONNECT_USER_ID": "",
"QUANTCONNECT_API_TOKEN": "",
"AGENT_NAME": "MCP Server"
}
}
}
}To get your user Id and API token, see [Request API Token](https://www.quantconnect.com/docs/v2/cloud-platform/com
fbbd823aad27OBSERVED · 2026-10-07Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add mcp-server --env QUANTCONNECT_API_TOKEN=${QUANTCONNECT_API_TOKEN} -- uvx mcp-server{
"mcpServers": {
"mcp-server": {
"command": "uvx",
"args": [
"mcp-server"
],
"env": {
"QUANTCONNECT_API_TOKEN": "${QUANTCONNECT_API_TOKEN}"
}
}
}
}Exposed tools (64)
34 read · 24 write · 6 destructive. Blast radius: 6 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
abort_optimization | read | Abort an optimization. |
authorize_connection | read | Authorize an external connection with a live brokerage or |
broadcast_live_command | read | Broadcast a live command to all live algorithms in an |
check_initialization_errors | write | Run a backtest for a few seconds to initialize the algorithm |
check_syntax | read | Check the syntax of a code. |
complete_code | read | Show the code completion for a specific text input. |
create_backtest | write | Create a new backtest request and get the backtest Id. |
create_compile | write | Asynchronously create a compile job request for a project. |
create_file | write | Add a file to a given project. |
create_live_algorithm | write | Create a live algorithm. |
create_live_command | write | Send a command to a live trading algorithm. |
create_optimization | write | Create an optimization with the specified parameters. |
create_project | write | Create a new project in your default organization. |
create_project_collaborator | write | Add a collaborator to a project. |
delete_backtest | destructive | Delete a backtest from a project. |
delete_file | destructive | Delete a file in a project. |
delete_object | destructive | Delete the Object Store file of a specific organization and |
delete_optimization | destructive | Delete an optimization. |
delete_project | destructive | Delete a project. |
delete_project_collaborator | destructive | Remove a collaborator from a project. |
enhance_error_message | read | Show additional context and suggestions for error messages. |
estimate_optimization_time | read | Estimate the execution time of an optimization with the |
liquidate_live_algorithm | write | Liquidate and stop a live algorithm. |
list_backtests | read | List all the backtests for the project. |
list_live_algorithms | read | List all your past and current live trading deployments. |
list_object_store_files | read | List the Object Store files under a specific directory in |
list_optimizations | read | List all the optimizations for a project. |
list_projects | read | List the details of all projects. |
lock_project_with_collaborators | write | Lock a project so you can edit it. |
patch_file | write | Apply a patch (unified diff) to a file in a project. |
read_account | read | Read the organization account status. |
read_backtest | read | Read the results of a backtest. |
read_backtest_chart | read | Read a chart from a backtest. |
read_backtest_insights | read | Read out the insights of a backtest. |
read_backtest_orders | read | Read out the orders of a backtest. |
read_compile | read | Read a compile packet job result. |
read_file | read | Read a file from a project, or all files in the project if |
read_latest_mcp_server_version | read | Returns the latest version of the QC MCP Server released. |
read_lean_versions | read | Returns a list of LEAN versions with basic information for |
read_live_algorithm | read | Read details of a live algorithm. |
read_live_chart | read | Read a chart from a live algorithm. |
read_live_insights | read | Read out the insights of a live algorithm. |
read_live_logs | read | Get the logs of a live algorithm. |
read_live_orders | read | Read out the orders of a live algorithm. |
read_live_portfolio | read | Read out the portfolio state of a live algorithm. |
read_mcp_server_version | read | Returns the version of the QC MCP Server that |
read_object_properties | read | Get Object Store properties of a specific organization and |
read_object_store_file_download_url | read | Get the URL for downloading files from the Object Store. |
read_object_store_file_job_id | write | Create a job to download files from the Object Store and |
read_optimization | read | Read an optimization. |
read_project | write | List the details of a project or a set of recent projects. |
read_project_collaborators | read | List all collaborators on a project. |
read_project_nodes | read | Read the available and selected nodes of a project. |
search_quantconnect | read | Search for content in QuantConnect. |
stop_live_algorithm | write | Stop a live algorithm. |
update_backtest | write | Update the name or note of a backtest. |
update_code_to_pep8 | write | Update Python code to follow PEP8 style. |
update_file_contents | write | Update the contents of a file. |
update_file_name | write | Update the name of a file. |
update_optimization | write | Update the name of an optimization. |
update_project | write | Update a project |
update_project_collaborator | write | Update collaborator information in a project. |
update_project_nodes | write | Update the active state of the given nodes to true. |
upload_object | write | Upload files to the Object Store. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (4 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (1)
delete_backtest, delete_file, delete_object, delete_optimization, delete_project, delete_project_collaborator
Gates applied: no_behavioural_pass.
fbbd823aad27full audit observations/trust-audit/mcp-server/quantconnect__quantconnect.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | fbbd823aad27 | SAFE | B | 89 | first audit |
Questions
What is the QuantConnect MCP server?
Official Python MCP server for local interactions with the QuantConnect API
What tools does QuantConnect expose?
64 in total: 34 read-only, 24 that write, and 6 that can delete or overwrite (delete_backtest, delete_file, delete_object, delete_optimization, delete_project). Every one is listed on this page with its risk.
Is QuantConnect safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 6 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does QuantConnect need?
It reads QUANTCONNECT_API_TOKEN from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How current is this page?
The grade is for one exact copy of the source (fbbd823aad27), read on 2026-10-07. The repository is watched and re-audited when it changes.