Skyvern
Type: full-code · Vendor: Skyvern-AI · Language: Python, TypeScript · License: AGPL-3.0 · Status: active · Status in practice: emerging
Automate browser-based workflows with AI using vision LLMs instead of brittle selectors.
Description. Playwright-compatible browser automation that uses vision LLMs plus computer vision to interact with websites without per-site XPath/CSS selectors, adapting to layout changes and sites it has never seen. Ships a no-code workflow builder alongside the SDK; reports strong WebVoyager and form-filling results.
Agent loop shape. A vision-LLM observes the rendered page, decides the next interaction, and acts via a Playwright-compatible layer — looping perceive→decide→act without relying on code-defined selectors.
Primary use cases
- automating browser workflows across unfamiliar or changing sites
- form-filling and data-entry automation without per-site selectors
- no-code browser automation for non-technical users
Key concepts
- Vision-LLM interaction → computer-use (docs) — The agent sees the rendered page and acts, instead of parsing the DOM.
- Workflow builder → visual-workflow-graph — No-code composition of multi-step browser automations.
Patterns this full-code implements —
- ★Computer Use
Automate Browser-based workflows using LLMs and Computer Vision
- ★Browser Agent
It provides a Playwright-compatible SDK that adds AI functionality on top of playwright
- ★★Visual Workflow Graph
a no-code workflow builder to help both technical and non-technical users automate manual workflows on any website
Neighbourhood
Click any neighbour to follow the lineage. Scroll to zoom, drag to pan.
Instantiates
Alternatives & relatives
Listed as alternative by (2)
References
Provenance
- Last analyzed:
- Last updated:
- Verification status: partial