shantycrawl-mcp
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@shantycrawl-mcpscrape https://example.com"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
shantycrawl-mcp
A lean MCP server for Firecrawl with lazy-loading tool architecture.
Replaces the official firecrawl-mcp (28 tools loaded upfront) with just 5 base tools always visible — advanced tools load only when you enable them via tool_enable. Drastically reduces context consumption in AI coding sessions.
Quick Start
# Run directly with npx — no install needed
FIRECRAWL_API_URL=http://localhost:3002 npx shantycrawl-mcpPoint FIRECRAWL_API_URL at your own Firecrawl instance (Docker, cloud, or self-hosted). Defaults to http://localhost:3002.
Related MCP server: webcrawl-mcp
Tools
Always Available (5)
Tool | Description |
| Extract markdown from a URL |
| Crawl a website for page content |
| Search the web |
| Load an advanced tool into the session |
| Unload an advanced tool from the session |
Lazy-Loaded via tool_enable (23)
tool_enable map adds map to your tool list for the session. Works for any of:
Discovery: map, extract, parse, check_crawl_status
AI Agent: agent, agent_status
Browser Interaction: interact, interact_stop
Feedback: search_feedback, feedback
Research (arXiv): research_search_papers, research_inspect_paper, research_read_paper, research_related_papers, research_search_github
Monitoring: monitor_create, monitor_check, monitor_checks, monitor_delete, monitor_get, monitor_list, monitor_run, monitor_update
Configuration
Claude Desktop
{
"mcpServers": {
"firecrawl": {
"command": "npx",
"args": ["shantycrawl-mcp"],
"env": {
"FIRECRAWL_API_URL": "http://localhost:3002"
}
}
}
}OpenCode
{
"mcp": {
"firecrawl": {
"type": "local",
"command": ["npx", "shantycrawl-mcp"],
"environment": {
"FIRECRAWL_API_URL": "http://localhost:3002"
}
}
}
}Environment Variables
Variable | Default | Description |
|
| Your Firecrawl API base URL |
Why shantycrawl-mcp?
The official firecrawl-mcp registers all 28 tool schemas upfront in every tools/list response. Each schema carries verbose descriptions and niche parameters. Over a session, this burns significant context tokens.
shantycrawl-mcp solves this with a simple pattern:
By default, only 5 essential tools appear — minimal schemas, minimal tokens.
Run
tool_enable <name>to activate any advanced tool for the session.The server emits
notifications/tools/list_changed, and the client sees only what you need.
Development
git clone https://github.com/shanty/shantycrawl-mcp.git
cd shantycrawl-mcp
npm install
npm run build
npm startArchitecture
Stdio transport — pure
@modelcontextprotocol/sdk, no FastMCPIn-memory state — tracks active tools per session
HTTP bridge — calls your Firecrawl API via native
fetchZero runtime dependencies beyond the MCP SDK
License
MIT
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Tools
Latest Blog Posts
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/schlemperdev/shantycrawl-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server