Frontrun MCP Server
# Frontrun MCP Server
VC follow intelligence for AI agents. Track what top investors follow on X — detect new follows, convergence signals, and trending companies before they're announced.
33 tools. Version 2.5.2.
## Setup (1 minute)
### Option A: OAuth login (recommended)
No API key needed. Log in with your frontrun.vc account:
```bash
# Step 1: Log in — opens browser, saves credentials locally
npx frontrun-mcp-server --login
# Step 2: Auto-configure your client
npx frontrun-mcp-server --setup # Claude Desktop
npx frontrun-mcp-server --setup-code # Claude Code (uses `claude mcp add`)
```
Credentials are saved to `~/.frontrun/credentials.json`. Other commands:
```bash
npx frontrun-mcp-server --status # Check auth status
npx frontrun-mcp-server --logout # Clear saved credentials
```
### Option B: Hosted endpoint (no install)
The server is also hosted at `https://www.frontrun.vc/mcp` (streamable HTTP). Point any
remote-capable MCP client at it with your API key:
```bash
claude mcp add --transport http frontrun https://www.frontrun.vc/mcp --header "Authorization: Bearer sig_your_key_here"
```
### Option C: Manual API key
If you prefer to use an API key directly, go to [frontrun.vc](https://frontrun.vc) → **Settings > API Keys** and generate a key (starts with `sig_`).
<details>
<summary><b>Claude Code</b></summary>
```bash
claude mcp add frontrun -e FRONTRUN_API_KEY=sig_your_key_here --scope user -- npx frontrun-mcp-server
```
Or add to `.mcp.json` in any project:
```json
{
"mcpServers": {
"frontrun": {
"command": "npx",
"args": ["frontrun-mcp-server"],
"env": {
"FRONTRUN_API_KEY": "sig_your_key_here"
}
}
}
}
```
</details>
<details>
<summary><b>Claude Desktop</b></summary>
Add to your config file:
- macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
- Windows: `%APPDATA%\Claude\claude_desktop_config.json`
```json
{
"mcpServers": {
"frontrun": {
"command": "npx",
"args": ["frontrun-mcp-server"],
"env": {
"FRONTRUN_API_KEY": "sig_your_key_here"
}
}
}
}
```
</details>
<details>
<summary><b>Cursor</b></summary>
Add to `.cursor/mcp.json` in your project:
```json
{
"mcpServers": {
"frontrun": {
"command": "npx",
"args": ["frontrun-mcp-server"],
"env": {
"FRONTRUN_API_KEY": "sig_your_key_here"
}
}
}
}
```
</details>
<details>
<summary><b>Any MCP-compatible client</b></summary>
```bash
FRONTRUN_API_KEY=sig_your_key_here npx frontrun-mcp-server
```
</details>
### Start asking questions
Open your AI tool and try:
> "What's trending in my tracked accounts this week?"
---
## What can I ask?
Here are real prompts that work out of the box. Just type them into Claude Code, Claude Desktop, or Cursor.
### Daily deal flow
- "What are the trending companies this week?"
- "Show me what my tracked VCs followed in the last 48 hours"
- "Any convergence signals? Who are multiple VCs following independently?"
### Deep research
- "Tell me everything about @hitdotone — who founded it, what they're building, and who's backing them"
- "Who are the founders of @someproject and what's their background?"
- "What funding has @tempofinance raised?"
### Discovery
- "Find me early-stage DeFi companies that my tracked VCs are following but I'm not tracking yet"
- "What sectors are getting the most attention from VCs right now?"
- "Search for stablecoin startups in the follow graph"
### Portfolio monitoring
- "Track @a16zcrypto and @paradigm"
- "Show me a16zcrypto's recent follow activity and what sectors they're focused on"
- "Find VCs with similar follow patterns to @multicoin"
### Custom intelligence
- "Create a rule to flag any stablecoin neobank companies"
- "Tag @hitdotone as a portfolio candidate"
- "Show me all entities matching my stablecoin rule from the last 7 days"
---
## How it works
Frontrun monitors the X (Twitter) follow graphs of VCs and investors you choose to track. When a VC follows a new account, Frontrun detects it, classifies the entity (startup, protocol, founder, etc.), and makes it queryable through AI.
**The signal:** VCs typically follow companies 2-8 weeks before a funding announcement. Convergence — multiple VCs independently following the same account — is the strongest signal.
**Receipts:** @techdollarhq flagged at 13 followers, $3M pre-seed announced 123 days later · @orthogonal_sh flagged 184 days before its $4.3M round led by Pantera · @rialto_xyz flagged at 18 followers, 26 days before its Robinhood partner announcement.
**What you're paying for:** Every response is computed intelligence, not raw data. Sectors are classified by AI, convergence is detected algorithmically, and company profiles are synthesized from multiple sources.
---
## Pricing
API + MCP access is included with the **Pro plan ($99/mo)** — it comes with 10,000 credits every month, and you can add more (or turn on auto-refill) at frontrun.vc/api/billing. Each query costs credits:
| What you're doing | Credits |
|---|---|
| Check your balance, list tracked accounts, manage rules/tags | Free |
| Preview an account, search, view sectors, pull a report | 4 |
| New follows, enriched follows, feed, company signals | 16 |
| Trending scan, VC activity | 24 |
| Thesis search, webhook setup (then 8/delivery) | 40 |
| Company deep-dive: overview, funding, convergence, resources, discover, similar VCs | 60 |
| Founder intelligence | 100 |
A typical daily check-in (feed + convergence + a few company deep-dives) runs ~300 credits — an included month covers it more than 30x over. Agent-scale usage is what auto-refill is for.
---
## All available tools
The agent has access to these tools automatically. You don't need to call them by name — just ask your question and the agent picks the right tool.
### Tracking
| Tool | What it does |
|---|---|
| `frontrun_status` | Your balance, tracked count, usage stats |
| `frontrun_list_tracked` | All accounts you're monitoring |
| `frontrun_track` | Start monitoring an X account |
| `frontrun_untrack` | Stop monitoring an account |
| `frontrun_preview` | Preview an account before tracking — signal score, sector hints, recommendation |
### Signals
| Tool | What it does |
|---|---|
| `frontrun_new_follows` | New follows detected across your tracked accounts |
| `frontrun_snapshot` | Current follow list for a specific tracked account |
| `frontrun_enriched_follows` | New follows + AI classification + custom rules applied |
| `frontrun_convergence` | Multiple tracked accounts following the same entity |
| `frontrun_trending` | Entities ranked by follow velocity |
| `frontrun_search` | Search discovered entities by sector, keyword, type |
| `frontrun_thesis_search` | Semantic search — describe a thesis in plain language, get matching companies |
| `frontrun_feed` | Real-time activity feed across all tracked accounts |
| `frontrun_sectors` | Sector breakdown of discovered entities |
| `frontrun_discover` | Personalized recommendations — "accounts your VCs follow that you're not tracking" |
| `frontrun_reports` | Generated intelligence reports for your tracked accounts |
### Company intelligence
| Tool | What it does |
|---|---|
| `frontrun_company` | Company overview — what they do, sector, stage |
| `frontrun_company_founders` | Founder profiles, backgrounds, previous companies |
| `frontrun_company_signals` | Social buzz, sentiment, notable engagements |
| `frontrun_company_resources` | Website, GitHub, docs, community links |
| `frontrun_company_funding` | Funding rounds, investors, amounts |
### VC intelligence
| Tool | What it does |
|---|---|
| `frontrun_vc_activity` | VC follow patterns — velocity, sector focus, recent follows |
| `frontrun_vc_similar` | Find VCs with overlapping follow patterns |
### Classification
| Tool | What it does |
|---|---|
| `frontrun_classify` | Run AI classification on specific entities |
| `frontrun_create_rule` | Create a custom classification rule |
| `frontrun_list_rules` | List your rules |
| `frontrun_update_rule` | Update a rule |
| `frontrun_delete_rule` | Delete a rule |
| `frontrun_tag` | Tag an entity with custom labels |
| `frontrun_list_tags` | List your tagged entities |
### Webhooks
| Tool | What it does |
|---|---|
| `frontrun_list_webhooks` | List your registered webhooks |
| `frontrun_create_webhook` | Register a webhook for new_follows / convergence events |
| `frontrun_delete_webhook` | Delete a webhook |
---
## REST API
Prefer HTTP? Every tool maps to a REST endpoint. Base URL: `https://frontrun.vc/v1`
```bash
# Set your key
export FR_KEY="sig_your_key_here"
# Check your account
curl -sL -H "X-API-Key: $FR_KEY" "https://frontrun.vc/v1/status"
# What's trending?
curl -sL -H "X-API-Key: $FR_KEY" "https://frontrun.vc/v1/trending?since=168h"
# Convergence signals
curl -sL -H "X-API-Key: $FR_KEY" "https://frontrun.vc/v1/convergence?since=168h"
```
Full REST docs: [frontrun.vc/docs/api](https://frontrun.vc/docs/api)
---
## Support
- Docs: [frontrun.vc/docs](https://frontrun.vc/docs)
- Issues: [github.com/jongall45/frontrun-mcp-server/issues](https://github.com/jongall45/frontrun-mcp-server/issues)
## License
MIT
TDQS
Scored across 18 tools
Every tool has a clearly distinct purpose with no ambiguity. For example, frontrun_account_activity analyzes a specific account's behavior, frontrun_convergence detects multi-account follow patterns, and frontrun_new_follows shows recent follows from tracked accounts. The descriptions clearly differentiate each tool's function, making misselection unlikely.
All tool names follow a consistent verb_noun pattern with the 'frontrun_' prefix, such as frontrun_create_rule, frontrun_list_tags, and frontrun_track. This uniformity makes the toolset predictable and easy to navigate, with no deviations in naming style.
With 18 tools, the count is slightly high but reasonable for the server's purpose of monitoring and analyzing social media follow activity. It covers tracking, classification, rule management, and data retrieval comprehensively, though it might be borderline heavy for some use cases.
The toolset provides complete CRUD/lifecycle coverage for the domain, including tracking (track/untrack), rule management (create/update/delete/list), data retrieval (account_activity, snapshot, new_follows), and analysis (convergence, trending, search). There are no obvious gaps that would hinder agent workflows.