Frequently Asked Questions about Firecrawl MCP Server on UBOS

Q: What is an MCP Server?

A: MCP stands for Model Context Protocol. An MCP server acts as a bridge, allowing AI models (specifically Large Language Models or LLMs) to access and interact with external data sources and tools. It standardizes how applications provide context to these LLMs.

Q: What is Firecrawl MCP Server?

A: The Firecrawl MCP Server is an implementation of the Model Context Protocol that integrates with the Firecrawl web scraping engine. This allows AI models to easily scrape, crawl, search, and extract data from the web.

Q: What are the key features of Firecrawl MCP Server?

A: Key features include: Web scraping with JS rendering, URL discovery and crawling, Web search with content extraction, Automatic retries with exponential backoff, Efficient batch processing, Credit usage monitoring, Support for cloud and self-hosted FireCrawl instances, and Mobile/Desktop viewport support.

Q: How do I install the Firecrawl MCP Server?

A: You can install the server using npm install -g firecrawl-mcp or by using npx -y firecrawl-mcp. For specific integrations with tools like Cursor, follow the configuration steps outlined in the documentation.

Q: What is Firecrawl API key and where can I find it?

A: The Firecrawl API key authenticates your requests to the Firecrawl service. You can find it in your Firecrawl account dashboard after signing up for their service.

Q: Is a Firecrawl API key required to use this?

A: Yes, if you’re using the cloud API. The FIRECRAWL_API_KEY environment variable is required. However, if you’re using a self-hosted Firecrawl instance, you can use the FIRECRAWL_API_URL environment variable instead (and the API key becomes optional depending on your self-hosted instance’s configuration).

Q: What environment variables can I configure?

A: Required: FIRECRAWL_API_KEY (for cloud API) or FIRECRAWL_API_URL (for self-hosted). Optional: FIRECRAWL_RETRY_MAX_ATTEMPTS, FIRECRAWL_RETRY_INITIAL_DELAY, FIRECRAWL_RETRY_MAX_DELAY, FIRECRAWL_RETRY_BACKOFF_FACTOR, FIRECRAWL_CREDIT_WARNING_THRESHOLD, FIRECRAWL_CREDIT_CRITICAL_THRESHOLD.

Q: What is the firecrawl_scrape tool?

A: This tool scrapes the content from a single URL, allowing various options like specifying output formats, extracting only the main content, setting a wait time for JavaScript rendering, and including or excluding HTML tags.

Q: What is the firecrawl_batch_scrape tool used for?

A: This tool efficiently scrapes content from multiple URLs simultaneously. It supports built-in rate limiting to prevent being blocked and offers parallel processing for faster scraping.

Q: How can I check the status of a batch scraping operation?

A: Use the firecrawl_check_batch_status tool and provide the ID of the batch operation.

Q: What does the firecrawl_search tool do?

A: The firecrawl_search tool performs a web search based on a given query and can optionally extract content from the search results.

Q: What’s the purpose of the firecrawl_crawl tool?

A: The firecrawl_crawl tool starts an asynchronous crawl of a website, allowing you to specify the maximum crawl depth, limit the number of pages, allow or disallow external links, and deduplicate similar URLs.

Q: How can I extract structured information from web pages?

A: Use the firecrawl_extract tool. This tool leverages LLMs to extract structured information based on a provided prompt, system prompt, and JSON schema. It supports both cloud-based and self-hosted LLMs.

Q: What kind of logging is included in the server?

A: The server includes comprehensive logging of operation status, performance metrics, credit usage, rate limit tracking, and error conditions.

Q: How does the server handle errors?

A: The server provides robust error handling, including automatic retries for transient errors, rate limit handling with backoff, detailed error messages, credit usage warnings, and network resilience.

Q: What is UBOS and how does Firecrawl MCP Server integrate with it?

A: UBOS is a Full-stack AI Agent Development Platform designed to help businesses orchestrate AI Agents, connect them with enterprise data, build custom AI Agents, and create Multi-Agent Systems. The Firecrawl MCP Server integrates seamlessly with UBOS, allowing AI agents built on the platform to easily access and process real-time web data.

Q: Where can I find more information about UBOS?

A: Visit the UBOS website: https://ubos.tech for more information.