#226 · Primary category: AI Coding Assistants
vibium
The verification layer for coding agents
Project last updated:08/28/26
GitHub Stars
2.9K
Forks
185
Contributors
9
License
Apache-2.0
Why we included this project
Coding agents that build or modify web interfaces share a real weakness: they often can't tell whether what they produced actually works. Vibium exists to give them that check. It installs as a skill, so an agent like Claude Code or Codex can open a URL, find interactive elements, click, fill forms, and capture screenshots through simple commands, with an MCP server and JavaScript, Python, and Java client libraries as alternatives. Elements are located by visible text, form label, placeholder, or ARIA role, so the agent never has to reason about fragile CSS selectors. The tool is a single ~10MB binary built on WebDriver BiDi, and the browser downloads itself on first run, which keeps setup minimal. Teams that want agents to verify their own front-end work before handing it back should find this concrete and useful.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
superpowers
An agentic skills framework & software development methodology that works.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
skills
Skills for Real Engineers. Straight from my .agents directory.
opencode
The open source coding agent.
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.