Build an agent tool
Give an agent bounded search and scrape tools, preserve citations, and control context size.
RESOURCE CENTER
Ship reliable web context faster.
Practical guidance for designing search, scrape, and crawl systems:from a first request to a monitored production pipeline.
Choose a path
Use the workflow closest to what you are building.
Create a knowledge base
Crawl approved domains, normalize content, and refresh only pages that changed.
Monitor a market
Extract stable fields from product, policy, and pricing pages on a schedule.
From prototype to production
Test extraction in the playground, copy a request from the docs, then use changelog updates to keep your integration current.
Interactive playground
Inspect the exact Markdown produced for a public URL before writing integration code.
Use-case blueprints
See concrete architectures for research, RAG ingestion, and competitive intelligence.
Product updates
Follow new extraction controls, reliability improvements, and API behavior changes.
Ready to give your product better context?
Run a real URL in the playground or review the request and response formats in the API docs.