The map for building software with AI.
Directories of tools, MCP servers, skills, and resources, a running series of field notes, and a plain-language dictionary — everything you need to actually get good at vibecoding.
Tools
A curated directory of AI coding assistants, agents, and editors — sorted by what they're actually good at.
Models
The top-tier LLMs coders are actually using — context windows, per-token pricing, and what each one is best at.
MCP Servers
The Model Context Protocol connectors that give agents access to real tools and data — filesystems, databases, GitHub, and more.
Skills
Packaged, reusable capabilities that extend what an agent can reliably do — document handling, design, testing, and more.
Resources
Guides, blogs, communities, and learning material for going deeper on vibecoding beyond this site.
Blog
Series-based articles on workflows, prompting patterns, and lessons from building real projects with AI.
Dictionary
Plain-language definitions for vibecoding concepts — from "context window" to "agentic loop."
Latest from the blog
View all →The State of AI Coding Tools in 2026
A snapshot of where terminal agents, IDE copilots, and browser builders stand today, and what's converging across all three …
- Prompting Patterns
Breaking Big Tasks Into Agent-Sized Steps
'Build me a dashboard' is a wish, not a task. A practical method for decomposing large asks into steps an AI agent can actually …
- Prompting Patterns
Context Is the Whole Game
Most disappointing AI output traces back to one cause: the agent didn't have the context a competent human would have had. Here's …
Featured tools
View all →Claude Code
Terminal AgentAnthropic's official CLI agent that reads, edits, and runs code directly in your terminal — built …
View toolCursor
Editor & IDEAn AI-first code editor forked from VS Code, with deep inline chat, multi-file edits, and an …
View toolGitHub Copilot
Editor & IDEThe original AI pair programmer — inline code suggestions, chat, and an agent mode available across …
View tool