Connectors · MCP
Give your AI a live market-data connection
ChartGaze is a remote Model Context Protocol (MCP) server. Instead of copying charts into a chat, a compatible AI client can discover ChartGaze tools and request market context, news, events, alerts, watchlists, and—when your account and trading mode allow—broker account actions. Your AI asks for a tool; ChartGaze checks the signed-in user's access before returning account data or performing an action.
Connector URL
Add this endpoint in an MCP client that supports remote Streamable HTTP connections:
https://api.chartgaze.live/appWhen the client asks for authentication, choose its sign-in option and complete ChartGaze's OAuth flow. This connects the client to your ChartGaze account; it does not grant permission beyond your account's access and trading-mode settings.
What happens after you connect?
- The AI client requests the ChartGaze tool catalogue and shows the tools it supports.
- You ask a question. The client may call relevant market-data tools, such as market context, historical context, news, or economic events.
- For account-specific tools, ChartGaze uses your authenticated account and checks permissions server-side. Trading controls and broker availability still apply to every request.
- Review any consequential action in your client and ChartGaze dashboard. Never share broker passwords or API secrets in an AI chat.
MCP connector or developer API?
Use the MCP endpoint above for a compatible AI client. Developers building their own integration can use the API base https://api.chartgaze.live and the API authentication and endpoints described in the docs; the MCP URL is not a substitute for an API token or SDK configuration.
Follow a client-specific walkthrough: Claude, ChatGPT, or Grok. Setup menus and availability can vary by client and account.