Skip to main content
Glama
cubrik MCP

cubrik MCP

by target1m

cubrik MCP

Use your cubrik account from Claude, ChatGPT, Claude Code, Codex, Cursor, VS Code, Gemini CLI and any app that speaks the Model Context Protocol: create, test and run AI trading agents on the brokers you connected, backtest ideas on historical candles, read your portfolio, results and Analytics, copy ready-made agents from the Library, and trade on your connected platforms.

https://mcp.cubrik.ai/mcp

Sign in with your cubrik account (OAuth) the first time the app calls it; there is no key to copy. It comes with cubrik's Intermediate plan and every plan above it; your plan's limits apply as they do in cubrik.

Install

Claude (claude.ai, desktop, phone) — Settings → Connectors → Add custom connector, paste the address.

Claude Code — as a plugin, with the skills:

/plugin marketplace add target1m/cubrik-mcp
/plugin install cubrik@cubrik-mcp

or just the server:

claude mcp add --transport http cubrik https://mcp.cubrik.ai/mcp

then /mcp to sign in.

ChatGPT — Settings → Apps → turn on developer mode, then create an app with the address (on the ChatGPT plans that allow custom apps). The version listed in ChatGPT's app directory uses https://mcp.cubrik.ai/chatgpt/mcp and places no orders.

Codex — codex plugin add target1m/cubrik-mcp, or add the server to ~/.codex/config.toml:

[mcp_servers.cubrik]
url = "https://mcp.cubrik.ai/mcp"

Cursor — Settings → MCP → Add, or this repository as a plugin (.cursor-plugin/).

VS Code — run “MCP: Add Server”, choose HTTP, paste the address.

Gemini CLI — gemini extensions install https://github.com/target1m/cubrik-mcp.

Related MCP server: dMoERA MCP Server

What it can do

Area

Tools

Account and portfolio

get_account, list_platforms, get_portfolio, get_trade_history, get_trading_results, connect_platform

Agents

list_agents, get_agent, get_agent_runs, list_agent_templates, create_agent, update_agent, start_agent, pause_agent, pause_all_agents, test_agent, run_agent_now, delete_agent

Backtests

run_backtest, get_backtest, check_rules_now

Library

search_library, get_library_agent, copy_library_agent

Analytics

get_analysis, get_advice, apply_advice

Markets

get_price, get_candles, list_instruments, get_news, get_economic_calendar, get_market_overview

Notifications and runs

list_notifications, mark_notifications_read, get_conversation

Skills

list_skills, get_skill, save_skill, delete_skill

Your brokers

each connected platform's own tools, as <platform>_<tool> — balances, positions, orders, stops, closing

In Claude and ChatGPT, backtests, agents, the portfolio, the Library and Analytics appear as interactive cards. Prompts: daily_brief, new_agent, backtest_idea, weekly_review, emergency_stop.

Skills

Skill

What it does

cubrik-setup

First steps: the account, the plan, connecting a platform

cubrik-agent-builder

An idea → questions → backtest → a paused agent → started on your word

cubrik-backtesting

Rules written right, run, and read honestly

cubrik-portfolio-check

Balances, positions, agents and what is coming up

cubrik-performance-review

Results, Analytics and advice you can apply

cubrik-library

Find, compare and copy ready-made agents

cubrik-trading

Orders, stops and closing positions, platform by platform

cubrik-skill-author

Write and save your own cubrik skills

Safety

  • You choose what each app may do when you sign it in: read, manage agents, trade.

  • No tool moves money into or out of any account.

  • Your broker keys stay with cubrik; the app gets a cubrik token for what you allowed.

  • Disconnect any app at any time: cubrik → Settings → AI apps.

Trading carries a risk of losing capital. cubrik is software that runs strategies you define; it does not give investment advice. Past results, backtests included, do not predict future ones.

Privacy · Terms · help@cubrik.ai

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      D
      maintenance
      Enables AI agents to trade crypto with paper money, access market data, view leaderboards, and manage trading bots via an MCP-compatible interface.
      16
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI agents to discover trading domains, inspect bots, backtest and submit crypto trading strategies for validation and live deployment, and monitor tournament and leaderboard status via MCP tools.
      9
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI agents to run a full quant research workflow over MCP: pulling data, authoring and backtesting strategies, running statistical validation and risk checks, and recording findings for future sessions.
      1
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      Enables AI agents to operate a single OctoBot instance through MCP, allowing them to manage trading profiles, run and compare backtests, and control portfolio, trading, exchange, and tentacle configuration surfaces via OctoBot's HTTP API.
      47
      MIT