MCP server - Bridgly

Give your AI client access to structured public data from Reddit, X (Twitter), LinkedIn, etc. The Bridgly MCP server makes the supported scraping API endpoints available as Model Context Protocol tools, so your client can fetch profiles, posts, discussions, search results, and more within a conversation or agent workflow. Connect Claude, ChatGPT, Cursor, Codex, or another compatible MCP client. The server uses Streamable HTTP at:

https://mcp.bridgly.app/mcp

There are two ways to authenticate:

Sign in with your Bridgly account. Clients that support OAuth, such as Claude and ChatGPT, only need the server URL. They open a Bridgly page where you sign in and approve access. See Sign in with your account.
Use an API key. Send your Bridgly API key (bgly_…) as a bearer token, just like the REST API. See Connect with an API key.

Scraping tools follow the same credit rules as their matching API endpoints and use the balance of the connected account. Checking your balance with billing_get_info does not consume credits. In your AI app, add https://mcp.bridgly.app/mcp as a custom connector, choose Connect, then sign in to Bridgly in the browser page that opens and allow access. See Sign in with your account for steps for each client, what a connected app can do, and how to disconnect one.

Connect with an API key

Need a key? Create one on the API Keys page of your dashboard. Keep the key private and out of shared configuration files — anyone with it can spend your credits.

Use an API key with clients that are configured through a file or a command line.

Claude Code
Cursor
Claude Desktop
Codex

Add the server from the CLI:

claude mcp add --transport http bridgly https://mcp.bridgly.app/mcp \
  --header "Authorization: Bearer bgly_YOUR_KEY"

Then start a session and ask it to fetch a public profile or search for posts — the Bridgly tools appear automatically.

Add the server to your mcp.json (Cursor Settings → MCP → Add):

{
  "mcpServers": {
    "bridgly": {
      "url": "https://mcp.bridgly.app/mcp",
      "headers": { "Authorization": "Bearer bgly_YOUR_KEY" }
    }
  }
}

Open Settings → Developer → Edit Config and add the server:

{
  "mcpServers": {
    "bridgly": {
      "url": "https://mcp.bridgly.app/mcp",
      "headers": { "Authorization": "Bearer bgly_YOUR_KEY" }
    }
  }
}

Restart Claude Desktop and the Bridgly tools show up in the tools menu.

Add the server to ~/.codex/config.toml:

[mcp_servers.bridgly]
url = "https://mcp.bridgly.app/mcp"
http_headers = { Authorization = "Bearer bgly_YOUR_KEY" }

Available tools

The tool catalogue covers Bridgly’s published scraping API across all supported platforms. Tools are generated from the API’s route registry, so newly supported endpoints become available without a separate MCP integration.

Profiles & accounts

Fetch public profiles, company details, and account activity using the tools available for each platform.

Posts & discussions

Retrieve posts, comments, reply threads, and community content for research and analysis.

Search & discovery

Find people, posts, jobs, companies, and more with the search tools each platform supports.

Tool names follow their endpoint paths: /x/get-user becomes x_get_user, and /linkedin/search/jobs becomes linkedin_search_jobs. Each tool exposes the same input fields and descriptions as its matching API endpoint, helping your client choose the right tool and supply its arguments. Each scraping tool’s description states what a call costs. In general, one call costs one credit. The account tool billing_get_info returns your current spendable credit balance and billing settings. It takes no arguments.

Browse the API Reference for the full platform coverage, available endpoints, and request and response schemas. Your MCP client’s tool list shows the tools currently available from the server.

How results come back

Successful calls return the matching API response as JSON text. Scraping results use the { timeMs, creditsCost, data } envelope: data contains the result, creditsCost reports the credits used, and timeMs reports the request duration. The billing_get_info tool returns the account’s billing snapshot instead. If a call fails, the server marks the tool result as an error and includes a message describing the problem, such as invalid arguments, an invalid API key, or insufficient credits. Your client can use that message to decide what to do next.

Security notes

Stateless by design

Each request is handled in isolation and bound only to the key you send. No session state is shared between callers.

Your key, your credits

Scraping calls use the credit balance of the account the key belongs to. Create a separate key for each integration so you can revoke one without interrupting clients that use other keys.

​Sign in from your AI app

​Connect with an API key

​Available tools

Profiles & accounts

Posts & discussions

Search & discovery

​How results come back

​Security notes

Stateless by design

Your key, your credits

Sign in from your AI app

Connect with an API key

Available tools

How results come back

Security notes