DOCUMENTATION
v1.0.22 — latestAutomate any
browser workflow
with AI.
Describe what you want in plain English. Skyvern opens a real browser, reads the page visually, and completes the task.
> Initializing browser session…
> Navigating to target URL… [OK]
> Analyzing page with vision… [OK]
> Planning actions…
→ ACTIONS_PLANNED: 4
> Executing: fill_form
> Executing: click_submit
> Task completed successfully
→ DURATION: 3.2s
> READY▊
01 — NO CODE
Automate without code
Use the dashboard to run tasks and build workflows visually.
02 — API
Build with the API
Integrate browser automation into your product with Python, TypeScript, or REST.
03 — SELF-HOST
Self-host
Run Skyvern on your own infrastructure with your own LLM keys.
ARCHITECTURE
How Skyvern works
The agent loop, Planner-Agent-Validator architecture, and what happens under the hood when you run a task.
↗REFERENCE
Core concepts
Tasks, workflows, blocks, runs, credentials, browser sessions, and engines.
↗USE CASES
Download invoices
Log into vendor portals, find invoices, download PDFs.
↗Fill forms at scale
Submit applications, registrations, and compliance forms.
↗Extract data
Pull structured data from any site without an API.
↗Healthcare portals
Extract patient demographics and billing from EHR portals.
↗E-commerce purchasing
Search products, add to cart, and complete purchases on any store.
↗Government forms
Navigate complex government websites to register accounts or file forms.
↗Insurance quotes
Fill multi-step quote forms across providers in any language.
↗Job applications
Auto-fill and submit applications on Lever, Greenhouse, and more.
↗●
Open source ·
