Pulse Fetch
Let your agent fetch clean structured page content from hard-to-scrape URLs using configurable cost-versus-speed strategies.
Overview
Pulse Fetch is an MCP server for the Idea phase that extracts clean structured web page content with optional anti-bot bypass via Firecrawl or Bright Data.
What is this MCP server?
- Extracts clean structured content from web pages with anti-bot bypass paths
- Optional FIRECRAWL_API_KEY and BRIGHTDATA_API_KEY for Firecrawl and Bright Data Web Unlocker
- OPTIMIZE_FOR choices cost or speed (default cost) for scraping strategy
- STRATEGY_CONFIG_PATH markdown file for custom scraping strategy (default /tmp/pulse-fetch/strategy.md)
- MCP_RESOURCE_STORAGE memory or filesystem; npm package @pulsemcp/pulse-fetch via npx stdio
- Package version 0.2.14
- OPTIMIZE_FOR defaults to cost with speed alternative
- Two optional unlocker integrations: Firecrawl and Bright Data
What problem does it solve?
Research agents choke on bot-protected or noisy pages, returning garbage that breaks competitor scans and validation notes.
Who is it for?
Indie builders who research competitors, docs, and landing pages inside Claude Code or Cursor and hit anti-bot or layout noise.
Skip if: Large scheduled crawls, sites you are not allowed to scrape, or teams that forbid paid unlocker services.
What do I get? / Deliverables
Your agent gets normalized page content from pulse-fetch using cost- or speed-tuned strategies and optional unlocker keys.
- Structured text content from target URLs in agent context
- Configurable scraping strategy via strategy markdown and cost/speed optimization
Recommended MCP Servers
Journey fit
Idea/research is the canonical shelf because competitive and market research is the first place solo builders hit bot walls. Research fits pulse-fetch’s core job: turn messy HTML into structured text your agent can cite, compare, or summarize.
How it compares
Focused fetch-and-extract MCP, not a full browser automation suite or hosted search API.
Common Questions / FAQ
Who is Pulse Fetch for?
Solo builders and agent users who need dependable single-page extraction during research, scoping, or content intel workflows.
When should I use Pulse Fetch?
When Idea or Validate research needs readable text from URLs that block simple fetchers, or when Launch SEO work needs structured competitor page copy.
How do I add Pulse Fetch to my agent?
Run npx @pulsemcp/pulse-fetch (stdio MCP), set FIRECRAWL_API_KEY and/or BRIGHTDATA_API_KEY if needed, and optionally OPTIMIZE_FOR and STRATEGY_CONFIG_PATH in your MCP config.