Frontend framework built around a View–Behavior–Logic architecture, designed to make application structure explicit and AI-assisted development reliable over time.
Enforce human-authored architectural invariants as a ratcheted gate, so AI-built codebases don't silently rot. Real-AST (Go/TS/JS/Python) + heuristic; pre-commit + AI-loop + debt-trend; god-object/coupling + layering + no-rotting-tests checks.
Ship AI-generated code to production with confidence - security, auth, and quality gaps caught before they ship.
An AI-native Markdown workspace where humans write and AI maintains knowledge.
Raw dataclasses + SQL for AI-generated code: zero ORM magic, slots=True, SQLite, 100% AI-readable
Deterministic code-planning tool. Transform flowcharts into Nassi-Shneiderman structograms, export to PDF/images (up to A0/B0), and generate code for 13 languages. Optionally connect API (cloud) or local (Ollama, LM Studio) LLM models for structured code generation.
Deterministic semantic CI that flags intent drift in AI-assisted code changes (experimental, pre-release)
Python → C++ converter for competitive programming. React + FastAPI + Ollama (Qwen). Side-by-side run, compile-repair loop, CF boilerplate.
Validate AI-generated CAD-as-code PRs (CadQuery / Build123d) — geometric metric diff, min-wall-thickness, DFM rules, and 6-view rendered PR previews. CLI + GitHub Action.
Find code that is BUILT but never WIRED. Reachability analysis from real entry points + a declared-boundary check. The CI gate for AI-generated code that compiles, passes tests, and is never called.
A resource-oriented systems language for AI-written code that should land on fast paths by default.
To associate your repository with the ai-codegen topic, visit your repo's landing page and select "manage topics."