# frontrun-mcp-server [Health: Active]

**Category:** 🌐 Social Media  
**Repository:** https://github.com/jongall45/frontrun-mcp-server  
**GitHub Stars:** 1  
**npm Downloads (last month):** 868  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/frontrun-mcp-server

## Description
Track what 1,000+ VCs follow on X to surface startups before their round. Signal for AI agents.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "frontrun-mcp-server": {
    "command": "npx",
    "args": ["-y","frontrun-mcp-server"],
    "env": {
      "FRONTRUN_API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `FRONTRUN_API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What frontrun-mcp-server does

The frontrun-mcp-server MCP server exposes Frontrun’s investor-follow data and intelligence to AI agents. It focuses on the X follow graphs of selected venture capital firms and investors. When tracked accounts follow new profiles, Frontrun makes those changes available for analysis, including entity classification, sector grouping, and overlap between investors.

Use it for deal-flow monitoring, early company discovery, portfolio research, and investor-pattern analysis. The data supports questions about trending entities, recent follows, funding, founders, company resources, social signals, and sectors. It also supports personalized discovery: an agent can identify accounts followed by tracked VCs that are not yet in your own tracking list.

## How it works

You first add X accounts to a tracked list. Frontrun then monitors their follow activity and provides the resulting data through MCP tools. The service can classify discovered entities as categories such as startups, protocols, or founders, and can apply rules you define. Convergence queries identify entities followed independently by multiple tracked accounts.

The server is designed for natural-language use. Instead of selecting a tool manually, you can ask an MCP client for recent activity, a company profile, or companies matching an investment thesis. The agent chooses among the available tools.

## Setup and configuration

The frontrun-mcp-server MCP server can run locally through npx or be accessed at `https://www.frontrun.vc/mcp` over streamable HTTP. For local use, install Node.js tooling capable of running `npx frontrun-mcp-server`.

OAuth is the recommended local setup. Run `npx frontrun-mcp-server --login` to open a browser and save credentials at `~/.frontrun/credentials.json`. The `--setup` option configures Claude Desktop, while `--setup-code` uses `claude mcp add` for Claude Code. `--status` checks authentication and `--logout` removes saved credentials.

API-key authentication is another option. Generate a key in Frontrun under Settings > API Keys, then provide it as `FRONTRUN_API_KEY` for a local server or as a Bearer header for the hosted endpoint. API and MCP access are included with the Pro plan, listed as $99 per month with 10,000 monthly credits. Tool calls consume different credit amounts depending on the operation.

## Tools and capabilities

The available tools cover five main areas:

- Tracking accounts, previewing accounts, and checking status or usage.
- Reviewing new follows, snapshots, enriched follows, convergence, trends, feeds, sectors, searches, thesis matches, recommendations, and reports.
- Researching companies, founders, signals, resources, and funding.
- Examining VC activity and finding investors with similar follow patterns.
- Managing classification rules, tags, and webhooks for new-follow or convergence events.

The same functionality is also exposed through REST endpoints under `https://frontrun.vc/v1`.

## Limitations and notes

Access is account-based and requires Frontrun authentication, either through saved OAuth credentials or an API key. The hosted endpoint also requires a remote-capable MCP client. Usage is metered with credits: account management operations are free, while research, signal, company, founder, thesis, and webhook operations have stated costs.

_Full upstream README: https://allmcps.com/mcp/frontrun-mcp-server/readme_