MCP Web Search & Scrape in MCPorter

Connect MCPorter, a TypeScript runtime and CLI for MCP servers, to Firecrawl through the hosted MCP server.

Quick setup

Create a Firecrawl API key and export it:

export FIRECRAWL_API_KEY="fc-YOUR-API-KEY"

Add Firecrawl to config/mcporter.json in your project, or to ~/.mcporter/mcporter.json to use it everywhere:

{
  "mcpServers": {
    "firecrawl": {
      "baseUrl": "https://mcp.firecrawl.dev/v2/mcp",
      "headers": { "Authorization": "Bearer ${FIRECRAWL_API_KEY}" }
    }
  }
}

MCPorter replaces ${FIRECRAWL_API_KEY} from your environment, so the key stays out of the file. Confirm the connection:

npx mcporter list firecrawl

Prefer a local process? Use the stdio form instead:

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": { "FIRECRAWL_API_KEY": "${FIRECRAWL_API_KEY}" }
    }
  }
}

Try it

npx mcporter call firecrawl.firecrawl_scrape url=https://firecrawl.dev 'formats=["markdown"]'

The command prints the page as markdown. Use the same entry from code with createRuntime (see the MCPorter docs).

Troubleshooting

Server not found: run npx mcporter config list to see which config file MCPorter loaded; --config <file> or MCPORTER_CONFIG selects one file only.
Authentication errors: MCPorter rejects the ${env:VAR} form; write the placeholder as ${FIRECRAWL_API_KEY} and check the variable is exported in the same shell.
Tool arguments are rejected: pass JSON values quoted for your shell, as in the formats example above.

Getting Started

AI Tools

Agent Harnesses

MCP Clients and Gateways

LLM SDKs and Frameworks

Agentic Debugging

Cookbooks

MCP Web Search & Scrape in MCPorter

Quick setup

Try it

Troubleshooting

​Quick setup

​Try it

​Troubleshooting

Quick setup

Try it

Troubleshooting