dalalos-mcp

DalalOS is a hosted Model Context Protocol (MCP) server that gives AI assistants sourced, filing-grounded Indian (NSE/BSE) stock-market data. Every response carries its source and freshness date. DalalOS returns sourced facts and mechanical measures, never buy/sell calls or target prices, and is not investment advice.

The server runs at https://mcp.dalalos.in/mcp (streamable HTTP, OAuth sign-in in the browser, free tier, no card). If your AI client accepts a server URL, connect it to that URL and sign in; you don't need this package at all. If your client can only launch a local command, dalalos-mcp proxy bridges it to the hosted server with an API key.

pip install dalalos-mcp
dalalos-mcp config claude-code   # print the connection snippet for a client
dalalos-mcp check                # confirm the hosted endpoint answers
dalalos-mcp tools                # list the tool names it advertises
dalalos-mcp proxy                # local stdio server for command-only clients (needs DALALOS_API_KEY)

Clients for config: claude, claude-code, codex, chatgpt, cursor, generic, stdio.

How sign-in works

• Clients that take a URL (Claude, ChatGPT, Claude Code, Codex, Cursor): the first time your client connects, DalalOS answers that sign-in is needed and your client opens the DalalOS sign-in page in your browser (email code or Google). Your DalalOS account is created on first sign-in, and your client keeps the login. You never copy a key.
• Clients that only launch a command: create an API key at https://dalalos.in/dashboard/keys and put it in the client's config as DALALOS_API_KEY (see "Command-only clients" below). The proxy sends the key only to mcp.dalalos.in.

Connect

Claude

Settings -> Connectors -> Add custom connector -> paste https://mcp.dalalos.in/mcp -> Connect, then sign in.

Claude Code

claude mcp add --transport http dalalos https://mcp.dalalos.in/mcp

ChatGPT

Install the published DalalOS plugin (India): https://chatgpt.com/plugins/plugin_asdk_app_6aded6de90f081918717da5227322e16 . Fallback: Plugins -> Create Plugin -> Server URL https://mcp.dalalos.in/mcp -> OAuth -> sign in.

Cursor and other MCP clients

{
  "mcpServers": {
    "dalalos": {
      "url": "https://mcp.dalalos.in/mcp"
    }
  }
}

Each client has its own schema for remote servers; see https://dalalos.in/connect.

Command-only clients (Claude Desktop config file, Cline and others)

pip install dalalos-mcp, create an API key at https://dalalos.in/dashboard/keys, then add:

{
  "mcpServers": {
    "dalalos": {
      "command": "dalalos-mcp",
      "args": ["proxy"],
      "env": { "DALALOS_API_KEY": "<your DalalOS API key>" }
    }
  }
}

If the client can't find dalalos-mcp, use the full path that pip show -f dalalos-mcp lists, or "command": "python", "args": ["-m", "dalalos_mcp", "proxy"]. Keep the key out of shared or committed config files.

Try it

Use DalalOS to fetch the latest available quote and company profile for Reliance Industries. Tell me which DalalOS tools you called, the source of the data, and its freshness date. Do not use web search.

Tools

DalalOS exposes 59 current MCP tools. Compatibility aliases and a ping helper are also visible on the server; the maintained reference is https://dalalos.in/tools and the machine-readable list is capabilities.json.

Full reference: https://dalalos.in/tools

Boundaries

• End-of-day data, not a live feed. Not a broker; no order placement.
• No investment advice, buy/sell calls or target prices.
• Not affiliated with NSE, BSE, SEBI, Anthropic, OpenAI or Cursor.

Links

• Website: https://dalalos.in
• Connection guide: https://dalalos.in/connect
• Data sources and freshness: https://dalalos.in/docs/data-sources-and-freshness
• Public repository (documentation and capability contract): https://github.com/BeyondNoww/DalalOS-Indian-Stock-Research
• Contact: atharva@dalalos.in