How it works

InferEdge MCP allows your AI tools and agents to access financial data from InferEdge's database and run analysis using InferEdge's analytics library.

The MCP help your AI agents to get validated data with minimal instructions from trusted sources that are validated and scrubbed. This brings accuracy, reliability and consistency to your analysis and should help you save time spent on methodology and fact checking.

Setup

While these instructions are Claude and Codex specific, you can follow corresponding steps for other clients too. You can plug the MCP + skills in any MCP-compatible client.

Claude Instructions

1. Go to Organization Settings → Connectors, add https://mcp.inferedge.com/mcp and name it "InferEdge MCP"
2. Allowing our domain: Go to Organization Settings → Capabilities → Code Execution and add this domain to allowed list - mcp.inferedge.com
3. Connect your account
   1. Go to Personal Settings → Connectors → Click Connect / Authorize
   2. Complete Authentication
   3. Grant access to all tools

Claude Code Instructions

1. Open terminal and enter the following command:

claude mcp add --transport http inferege https://mcp.inferedge.com/mcp

2. Authenticate after starting claude on terminal by using /mcp command. Select the mcp server and follow the instructions that claude code will give you.

Codex Instructions

1. Go to Settings → MCP Servers, and click Add Server
   1. Give server name "inferedge_mcp" or any other name you like
   2. Select Streamable HTTP and add https://mcp.inferedge.com/mcp under URL
   3. Save - this will add the MCP server
2. To authenticate, open Terminal and enter the following command. Unfortunately, codex does not support authentication from the desktop app, you have to authenticate only via terminal.
   Shell

   codex mcp login <mcp_name_from_above>

3. You will have to restart codex desktop app after you authenticate on terminal.

Examples

You can now test whether the connection is working. Few sample questions to get you started:

• Get the recent social media posts on Nvidia
• Get comparison of analyst estimate and actual earnings for US consumer stocks

Other examples use cases include analysis around earnings guidance, searching filings and footnotes, impact of news on stocks or sectors, and real-time prices.