mcpField
G
Gorilla MCP

Gorilla MCP

Local
github.com/opusforgeBusiness & Marketing3Updated 15w ago

About Gorilla MCP

Gorilla MCP is a lead-discovery server for solo founders looking for early SaaS customers. It searches social platforms for expressed demand and ranks results by buying intent.

Key features

  1. 01Parallel searches across five platforms
  2. 02Hot, Warm, Cold intent scoring
  3. 03Retrieve and reread search results
  4. 04Check plan and credit balance

Use cases

  1. 01Find people seeking a product category
  2. 02Review results after a search times out
  3. 03Check remaining search credits

MCP support

Exposes search, result-retrieval, and billing-status tools to MCP clients. Searches Reddit, X, YouTube, LinkedIn, and Bluesky through the Gorilla API.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"gorilla": {
"command": "npx",
"args": [
"-y",
"github:opusforge/gorilla-mcp"
],
"env": {
"GORILLA_API_KEY": "<GORILLA_API_KEY>"
}
}
}
}
Run commandnpx -y github:opusforge/gorilla-mcp
Environment variables to set
GORILLA_API_KEYSecret

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmgithub:opusforge/gorilla-mcp

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

Gorilla MCP Server

CI Version License: MIT smithery badge gorilla-mcp MCP server

Stars Last commit Open issues TypeScript Node MCP SDK Built for Claude

Model Context Protocol server for Gorilla. Find your first 100 SaaS users by searching Reddit, X, YouTube, LinkedIn, and Bluesky for real demand signals. Start free with 100 credits. You spend one credit per qualified lead (hot or warm); low-relevance results are free.

Learn more: usegorilla.app · Real run examples · How Gorilla compares · Blog

gorilla-mcp MCP server

Setup

Run directly from GitHub with npx:

GORILLA_API_KEY=grla_... npx -y @usegorilla/mcp

Or clone and run locally:

git clone https://github.com/opusforge/gorilla-mcp
cd gorilla-mcp
npm install && npm run build
GORILLA_API_KEY=grla_... node dist/index.js
Get your API key
  1. Sign up at usegorilla.app. Free tier: 100 credits, no card.
  2. Sign in at platform.usegorilla.app.
  3. Menu, API Keys, Create.
  4. Copy the key (shown once).
Configure in Claude Code

Add to your Claude Code settings:

{
  "mcpServers": {
    "gorilla": {
      "command": "npx",
      "args": ["-y", "@usegorilla/mcp"],
      "env": {
        "GORILLA_API_KEY": "grla_your_key_here"
      }
    }
  }
}
Configure in Cursor

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "gorilla": {
      "command": "npx",
      "args": ["-y", "@usegorilla/mcp"],
      "env": {
        "GORILLA_API_KEY": "grla_your_key_here"
      }
    }
  }
}

Tools

search

Search Reddit, X, YouTube, LinkedIn, and Bluesky in parallel for posts where people express demand for what you describe. Returns results ranked Hot / Warm / Cold by buying intent. Takes 30-90 seconds. One credit per qualified lead (Hot or Warm); Cold results are free, and a failed search is refunded. LinkedIn is a paid-plan source; the free tier covers Reddit, X, YouTube, and Bluesky.

Parameters:

  • query (required) - what to search for, in your own words
  • source (optional) - reddit, twitter, youtube, linkedin, bluesky, or all (default)
  • since (optional) - 24h | 7d | 30d | all, or an ISO date. Default 7d.
  • limit (optional) - max results, 1-200. Default 50.

Returns: A search_id plus scored results (source, channel, title, URL, score, Hot/Warm/Cold) and the credits charged.


get_search

Fetch the current state and results for a search by its search_id. Use it to recover a search that timed out client-side, or to re-read a recent one. Free.

Parameters: search_id (required)


billing_status

Check your plan and remaining credit balance. Free.

Returns: Plan (free or monthly) and credit balance (tier + pack = total).

Example workflow

1. search("a language learning app for travelers")
   -> search_id + scored leads across Reddit, X, YouTube, LinkedIn, Bluesky,
      ranked Hot / Warm / Cold by buying intent. One credit per Hot/Warm lead.

2. get_search(search_id)   -> re-read or recover results without searching again

3. billing_status()        -> plan + credits remaining
Install via Smithery

Available at smithery.ai/server/opusforge/gorilla-mcp. Smithery distributes a pre-built MCPB bundle that any MCPB-compatible client can install in one click.

Environment variables

VariableRequiredDescription
GORILLA_API_KEYYesYour API key (starts with grla_). Create one at platform.usegorilla.app, Menu → API Keys.
GORILLA_API_BASENoOverride the API base. Default https://usegorilla.app/v1. Useful for self-hosted deployments.

No other configuration is required — the package talks to the public API with your key.

Pricing

  • Free tier: 100 credits, granted once, no card. A one-time trial.
  • Paid plan: $14.99/mo for 2,000 credits. Unused credits roll over.
  • Metering: one credit per qualified lead (hot or warm). Low-relevance results are free. Failed searches refund.
  • Sources: the free tier covers Reddit, X, YouTube, and Bluesky. The paid plan adds LinkedIn (all five).

search spends one credit per qualified lead it returns; Cold results are free. get_search and billing_status are free.

See usegorilla.app for the full product.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.