Back to Blog
Talk to Your Trading Bots: Introducing the Sigrex MCP Server
MCP4 min read

Talk to Your Trading Bots: Introducing the Sigrex MCP Server

Building trading automation usually means a lot of clicking. Open a strategy, edit a value, save, check the logs, switch to another bot, repeat. It works, but it's slow, and it pulls your attention away from the part that actually matters: the trading ideas themselves.

Today that changes. With the Sigrex MCP server, you can connect AI assistants like Claude and ChatGPT directly to your Sigrex account and manage your whole trading setup in plain language.

Ask a question, get an answer from your real account data. Describe a strategy, and your assistant builds it. Tell it to pause your bots, and it's done.


What is MCP?

MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external tools and services. Instead of copying and pasting data between apps, your assistant talks to Sigrex directly: it can read your strategies, bots and webhooks, and take actions on your behalf.

Think of it as giving your AI assistant a secure key to your Sigrex dashboard.

What's new

The Sigrex MCP server now uses OAuth sign-in and gives your assistant access to your entire account. There are no API keys to copy. You connect, sign in with your Sigrex account, approve access, and you're ready.

Everything you can do in the Sigrex app, your assistant can now do too:

  • Strategies: create, edit, duplicate, activate and pause code-based and AI-powered strategies
  • Trading bots: manage bots on crypto exchanges, DEXs and prediction markets
  • Webhooks: create webhooks and send trading signals
  • AI trading agents: build LLM sessions, change their prompts and models, and inspect their memory
  • Market data: check live prices across supported exchanges
  • Documentation: search the Sigrex docs without leaving the chat

Getting started in two minutes

  1. In your AI assistant, add a new custom connector (or MCP server) with this URL:
   https://mcp.sigrex.io
  1. Click Connect. You'll be sent to the Sigrex sign-in page.
  2. Sign in and approve the access request.
  3. That's it. Your assistant can now see and manage your Sigrex account.

Using VS Code, Cursor or another developer tool? Add it to your MCP config:

{
  "mcp": {
    "servers": {
      "sigrex.io": {
        "type": "http",
        "url": "https://mcp.sigrex.io"
      }
    }
  }
}

The full setup guide is in our documentation.


What you can do with it

Here are some real things you can ask, and why they're useful.

1. Get an instant overview of your setup

"Show me all my active strategies and when each one last triggered."

No more clicking through folders. In seconds you get a clear summary of what's running, what fired recently, and what's been quiet. Great as a morning check-in.

2. Build a strategy by describing it

"Create a strategy for ETHUSDC that goes long when RSI drops below 30 and the price is above the 1-hour EMA, and exits at 1% profit."

Your assistant writes the strategy code, creates it in your account and connects it to your webhook. You review it, adjust what you want, and activate it when you're ready. Going from idea to working strategy now takes minutes, not an afternoon.

3. Scale a strategy to new markets

"Duplicate my BTC momentum strategy for SOL, AVAX and LINK."

Found something that works on one pair? Your assistant copies it, updates the symbols and keeps everything else the same. Testing an idea across ten markets becomes a single sentence.

4. Debug a strategy that isn't behaving

"My POPCAT strategy hasn't entered a trade in two days. Look at its code and stored state and tell me why."

Your assistant reads the strategy code and its saved storage, walks through the entry conditions, and explains which filter is blocking trades. Instead of guessing, you get a clear answer, and often a suggested fix.

5. Build and supervise AI trading agents

"Create an AI session that checks the BTC 4-hour market every 4 hours, researches the direction and sends a signal to my prediction-market webhook."

Sigrex LLM sessions are AI agents that run on a schedule with live market data. With MCP, you can design them conversationally, then come back later and ask:

"Review the last 20 decisions of my BTC 4h session. How often was it right, and what patterns do you see in its mistakes?"

Your assistant reads the session's memory, analyzes its track record and helps you improve the prompt. It's like having a research partner for your AI agents.

6. Take control quickly

"Pause all my trading bots in the Xy folder."

When markets turn wild or you're stepping away, you don't need to open the app and toggle bots one by one. One message handles it.

7. Check the market on the fly

"What's the current price of BTC on Binance and Kraken?"

Live prices, right in your conversation, without switching tabs.


Why this matters

Trading automation has always had two layers: the ideas (what to trade and why) and the operations (building, configuring, monitoring and fixing). Most people spend far more time on operations than on ideas.

The Sigrex MCP server shifts that balance:

  • Faster iteration. Test more ideas in less time, because building and adjusting strategies happens in conversation.
  • Better understanding. Your assistant can read your code and your bots' history and explain what's actually happening.
  • Less friction. Routine management (overviews, pausing, duplicating, renaming) becomes a sentence instead of a dozen clicks.
  • Smarter AI agents. Use one AI to review and improve another, with real performance data.

You stay in charge of the decisions. Your assistant handles the busywork.


A note on safety

Because the MCP server has full access to your account, your assistant can take real actions, including starting strategies, sending live trading signals and deleting resources. A few good habits:

  • Review before you approve. Most AI assistants ask for confirmation before running an action. Read what it's about to do, especially for anything that trades or deletes.
  • Test before going live. Let your assistant build a strategy, then review the code yourself before activating it with real funds.
  • Only connect clients you trust. You can disconnect at any time by removing the connector from your assistant.

Automation is powerful, and so is AI. Used together carefully, they can save you a lot of time. Neither one guarantees profits, so keep your risk management in place.


Try it today

The Sigrex MCP server is available now for all Sigrex accounts.

  1. Create a free Sigrex account if you don't have one yet.
  2. Connect https://mcp.sigrex.io to your AI assistant.
  3. Ask it: "What can you do with my Sigrex account?"

We'd love to hear what you build with it. Share your favorite prompts and strategies with us, and check the documentation for the full setup guide.

Happy building.

Related posts

How to Forward Test a Strategy
Testing·5 min read

How to Forward Test a Strategy

Sigrex has no backtester, and that is a design consequence, not a missing feature. A strategy here can read any API and ask any LLM, and you can't replay that from history. What you can do is run the strategy on the live market with nothing attached to it, and watch. Why a backtester doesn't fit Sigrex A backtest replays the past. It feeds old data into your rules and counts what would have happened. That works when two things are true: every input the strategy uses has a recorded history, an

2026-10-02
Building a News-To-Trade Workflow for Polymarket
LLM Reaction·7 min read

Building a News-To-Trade Workflow for Polymarket

Prediction markets move on information. A market resolves on a specific, checkable question — did the rate cut happen, did the bill pass, did the company hit the number — and every time new information about that question becomes public, there's a brief window before the market price catches up with it. Most of that window closes in minutes. Nobody is checking Polymarket by hand fast enough to consistently catch it. This post walks through a pipeline that does the watching for you: Perigon deli

2026-08-26
Building Copy Trading Pipeline for Polymarket
Code Strategy·7 min read

Building Copy Trading Pipeline for Polymarket

Copy trading a wallet on Polymarket is straightforward in principle: every trade a wallet makes is public, on-chain, and available through a free API. The harder question is how to relay those trades into your own bot without inheriting the source wallet's exact position sizing, without double-processing the same fill twice, and without losing visibility into what actually got copied and why. This walks through one way to do that: a Code Strategy that polls a wallet, a Data Webhook that stores

2026-08-17