Cortex — Universal AI Context Engine

Define .cortex/ once — Cortex compiles your context, rules, and model preferences to Claude Code, Cursor, Copilot, Windsurf, Gemini CLI, and 4 more. One source of truth for your entire AI toolchain.

⚡

Stop rewriting rules

One .cortex/ directory. Nine provider formats generated automatically. No more copy-pasting .cursorrules across tools.

📊

Token budget, visible

Every rule is impact-scored and token-counted. Cortex trims, compresses, and warns before your AI hits a context wall mid-session.

🧠

Model-aware by default

Detects Claude, GPT, Gemini, Llama. Tunes context delivery to each model's strengths. Switch models without losing your project's memory.

🌐

Community intelligence

Import battle-tested rule sets for React, FastAPI, Go, and more. Your .cortex/ is portable, shareable, and git-tracked.

Works with Claude Code · Cursor · GitHub Copilot · Windsurf · Gemini CLI · OpenAI Codex · Kiro

Everything you need, one CLI

Install once globally or run via npx — zero config

$ npx cortex init Bootstrap .cortex/ in any project — detects your stack automatically

$ npx cortex compile Compile rules to all 9 AI provider formats at once

$ npx cortex score Analyze token budget, rule impact, and compression savings

$ npx cortex health Check model compatibility and provider health scores

$ npx cortex add react Install community rule packages for your stack

$ cortex switch claude Switch active AI provider and recompile context instantly

Stack Intelligence

Discover what you can build — or let us guide you to the right stack

Live Insights

Answer questions to see real-time architecture insights

What You Can Build	Core Stack	Complexity	Real Example	Match

.cortex/main.md