OSS TanbouOn this page
Overview
Skyvern combines LLM-based agents with Playwright to understand web pages, plan actions, and execute browser workflows. Its design avoids depending on large sets of fixed XPath or CSS selectors, aiming to handle unfamiliar sites and layout changes more robustly.
Features and best fit
Based on official documentation; not hands-on tested · Content checked:
Key features
Reason about pages and plan browser actions
Skyvern agents interpret websites and map visible elements to the actions needed for a workflow, rather than requiring every target site to be encoded in advance.
Sources: [2]
Reuse workflow intent across changing or different sites
Because workflows are not just lists of fixed selectors, the same automation can be applied to multiple sites and can be less brittle when layouts change.
Sources: [2]
Evaluate locally with pip or a containerized stack
The pip quick start can use SQLite by default, while Docker Compose starts the UI, API, and PostgreSQL together for a containerized setup.
Sources: [2]
Best fit
Fits AI automation of external SaaS and business portals
It is relevant for form entry, information gathering, and repetitive browser operations where the target system does not expose a sufficient API.
Sources: [2]
Before adoption
Treat LLM credentials and action validation as production controls
Local deployment needs an LLM provider configuration. For actions with material side effects, add confirmation, validation, retries, and auditability around model-driven browser actions.
Sources: [2]
CAPTCHA and anti-bot behavior can differ by deployment
The README specifically bundles anti-bot detection, proxy networking, and CAPTCHA solvers with the managed cloud product. Do not assume an identical environment in a self-hosted deployment.
Sources: [2]
Review AGPL-3.0 obligations for your delivery model
Skyvern is licensed under AGPL-3.0. Review source-availability obligations for modified deployments, including network-service scenarios, before production use.
Sources: [4]
Official sources
- [1]Skyvern-AI/skyvern repository(2026-10-01)
- [2]Skyvern README(2026-10-01)
- [3]Skyvern v1.0.54 release(2026-10-01)
- [4]Skyvern AGPL-3.0 license(2026-10-01)
Supplemental curator note
Skyvern is a browser agent that reasons about pages instead of encoding large sets of XPath or CSS selectors. That can reduce brittleness across changing sites, but production workflows still need validation, retries, and controls around LLM-driven actions.
Try it in 3 steps
- 1
Clone Skyvern
Fetch the official repository to evaluate the containerized UI, API, and PostgreSQL stack.
git clone https://github.com/skyvern-ai/skyvern.git && cd skyvern - 2
Configure the LLM provider
Set the chosen LLM provider API key and related values in .env. Do not commit credentials to the repository.
cp .env.example .env - 3
Start the container stack
Start Skyvern's API, UI, and PostgreSQL services using the Docker Compose path documented in the README.
docker compose up -d
Growth
Growth trends · Last 30 days
23,123 Stars
Trend data is still being collected.
Development activity
Last 90 days · weekly
- Commits (last 30 days)
- 275
- Open PRs
- 224
Development activity is still being collected.
Built with
Categories and tags
GitHub data
GitHub dataView detailed GitHub data
GitHub Topics
- api
- automation
- browser
- computer
- gpt
- llm
- playwright
- python
- rpa
- vision
- workflow
- browser-automation
Related information
Write a related articleShare a guide or use case for this OSS in Markdown. Articles are published after administrator approval.
Explore next
- Suna / Kortix20,240 Stars
3 shared tag(s) · 2 shared category(s)
an AI operations platform that versions agents, skills, memory, and connectors in Git and sends isolated-sandbox work through human-approved change requests
TypeScript - LlamaIndex52,383 Stars
3 shared tag(s) · 1 shared category(s) · Same language
compose ingestion, retrieval, RAG, and agents from modular packages
Python - TestZeus Hercules1,174 Stars
6 shared tag(s) · 2 shared category(s) · Same language
turn Gherkin scenarios into LLM-agent browser actions for end-to-end testing
Python - Dify157,686 Stars
5 shared tag(s) · 2 shared category(s)
design and publish AI workflows, retrieval, and agents on one platform
TypeScript - Keep12,373 Stars
5 shared tag(s) · 1 shared category(s) · Same language
centralize monitoring alerts and connect correlation, enrichment, and response workflows
Python - Browser Use116,992 Stars
4 shared tag(s) · 3 shared category(s) · Same language
give LLM agents real browser control through a Python library, agent CLI, or hosted browser infrastructure
Python
Report incorrect information
Tell us if any listing information is incorrect or outdated.