Features
Nine-stage AI code generation pipeline (change through integration)
Human approval gates at change, design, and review stages
Token compression for long session continuity
ChangeGraph dependency tracking
Multi-terminal support: Claude, Codex, Cursor, OpenCode
10 configurable workflow profiles (full, lite, nano)
TDD constraint (tests-first, red-green) in dev stage
Evidence-based review with executable verification per acceptance criterion
Project-level planning from README/PRD/PDF/URL
Auto mode full-stack skeleton generation
Built-in translation API service integration
Docker support for containerized workflows
Testing framework integration (Vitest)
Environment configuration via .env
Prompt management system with skills/agents framework
@ai_fn decorator for LLM-powered Python functions
@ai_classifier decorator for text classification
Structured data extraction via Pydantic models
Agent loops with tool calling and function calling
Built-in streaming (SSE) support
Async-first API for concurrent applications
Rate limiting and automatic retries
LLM response caching for performance
Concurrency control for workload management
CLI tool for monitoring and debugging
SQLite state store for persistence
OpenAI and Anthropic model support
Embeddings generation for semantic search
Local execution with no cloud dependency
Self-hosted deployable as a library
