Web UI

The web UI runs at http://localhost:3000 after starting npm run server.

Screens

Monitor

Real-time view of whatever job is running (or last ran).

• Action ticker — shows the agent's current action. Pulsing green dot = running, dim = idle/done.
• Stats bar — iteration count, scraped pages, records found, elapsed time, token usage
• Active Scrapes — URLs currently being fetched
• Search Queries — Google searches the agent has made
• Updated Rows — rows refreshed by an update job (green = changed)
• Completed Scrapes — finished scrapes with char counts
• Extractions — records extracted per page
• Jobs — last 5 jobs, click to load any into the monitor view
• Errors — any tool errors during the run
• Raw Log — full event stream

Scrape

Forms and data management.

• Index tab — start a new index job (topic, schema, output, iterations, seed URLs)
• Update tab — start an update job (dataset, schema, optional filter)
• Datasets — list of all datasets, with open/download/delete actions
• Schemas — create and edit schemas

Token counter

The stats bar shows a running total of tokens used during the current job. Hover to see the input/output split. Useful for estimating costs mid-run.

Settings

Click the gear icon in the header to configure:

• Crawl4AI endpoint — URL of your Crawl4AI instance (overrides the CRAWL4AI_BASE env var, takes effect immediately)
• LLM provider — Anthropic / OpenAI / ZordMind
• Agent model and extract model per provider
• API key — read-only display with copy button; required to trigger jobs via the REST API
• Allowed origins — comma-separated list of origins allowed for CORS (for Webflow or other external integrations)
• Webhook URL — called when a job finishes (POST with job result)