llms.txt Content
# Haunt API
> Agent-readable index for Haunt API. Prefer `Accept: text/markdown` on public pages.
## Core pages
- [Homepage](https://hauntapi.com/) - product overview, pricing, signup, examples.
- [Pricing](https://hauntapi.com/pricing.md) - agent-readable pricing: plans, monthly credits, rate limits, and per-extraction credit costs.
- [Docs](https://hauntapi.com/docs) - API usage, authentication, JSON/Markdown response shapes, rate limits.
- [Demo](https://hauntapi.com/#extractor) - live no-key demo on the homepage; fixed sample JSON, plus first-call curl after signup.
- [Agents](https://hauntapi.com/agents) - one-command CLI/MCP setup and agent skill curl for AI tools.
- [Agent skill](https://hauntapi.com/agents/skill.md) - compact setup instructions for agents to fetch with curl.
- [AI agent web extraction](https://hauntapi.com/use-cases/ai-agent-web-extraction) - focused no-key demo to free-key to first-extraction path.
- [Turn web page into JSON API](https://hauntapi.com/use-cases/turn-web-page-into-json-api) - URL-plus-prompt to structured JSON with proof and boundaries.
- [Extract pricing table from website](https://hauntapi.com/use-cases/extract-pricing-table-from-website) - visible plan cards and pricing tables to comparable JSON fields.
- [MCP server for web scraping](https://hauntapi.com/use-cases/mcp-server-for-web-scraping) - hosted and local MCP setup with try_demo_extract first.
- [What Haunt can extract](https://hauntapi.com/what-haunt-can-extract) - benchmark-backed capability proof and honest platform limits.
- [Recipes](https://hauntapi.com/recipes) - copy-paste extraction recipes for company pages, docs, GitHub, Reddit, and HN.
- [Blog](https://hauntapi.com/blog) - guides and extraction examples.
- [Privacy](https://hauntapi.com/privacy) - account, request, submitted-content, retention, and rights policy.
- [Terms](https://hauntapi.com/terms) - service rules, acceptable use, credits, billing, liability, and governing law.
- [Cookies](https://ha
OpenAPI Spec (preview)
{"openapi":"3.1.0","info":{"title":"Haunt API","description":"Extract structured data from permitted web pages with one API call. Supports JavaScript rendering, fallback fetch paths, and clear CAPTCHA-aware failures without bypassing human verification.","version":"1.0.0"},"paths":{"/v1/extract":{"post":{"tags":["Extraction"],"summary":"Extract Data","description":"Extract structured data from a permitted public URL using AI.\n\nSend a permitted public URL and describe what data you want in plai