forked from cardosofelipe/fast-next-template
Felipe Cardoso
88cf4e0abc
## Model Stack Updates (User's Actual Models) Updated all documentation to reflect production models: - Claude Opus 4.5 (primary reasoning) - GPT 5.1 Codex max (code generation specialist) - Gemini 3 Pro/Flash (multimodal, fast inference) - Qwen3-235B (cost-effective, self-hostable) - DeepSeek V3.2 (self-hosted, open weights) ### Files Updated: - ADR-004: Full model groups, failover chains, cost tables - ADR-007: Code example with correct model identifiers - ADR-012: Cost tracking with new model prices - ARCHITECTURE.md: Model groups, failover diagram - IMPLEMENTATION_ROADMAP.md: External services list ## Architecture Diagram Updates - Added LangGraph Runtime to orchestration layer - Added technology labels (Type-Instance, transitions) ## Self-Hostability Table Expanded Added entries for: - LangGraph (MIT) - transitions (MIT) - DeepSeek V3.2 (MIT) - Qwen3-235B (Apache 2.0) ## Metric Alignments - Response time: Split into API (<200ms) and Agent (<10s/<60s) - Cost per project: Adjusted to $100/sprint for Opus 4.5 pricing - Added concurrent projects (10+) and agents (50+) metrics ## Infrastructure Updates - Celery workers: 4-8 instances (was 2-4) across 4 queues - MCP servers: Clarified Phase 2 + Phase 5 deployment - Sync interval: Clarified 60s fallback + 15min reconciliation 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Syndarix
Your AI-Powered Software Consulting Agency
An autonomous platform that orchestrates specialized AI agents to deliver complete software solutions with minimal human intervention.
Vision
Syndarix transforms the software development lifecycle by providing a virtual consulting team of AI agents that collaboratively plan, design, implement, test, and deliver complete software solutions.
The Problem: Even with AI coding assistants, developers spend as much time managing AI as doing the work themselves. Context switching, babysitting, and knowledge fragmentation limit productivity.
The Solution: A structured, autonomous agency where specialized AI agents handle different roles (Product Owner, Architect, Engineers, QA, etc.) with proper workflows, reviews, and quality gates.
Key Features
Multi-Agent Orchestration
- Configurable agent types with base model, failover, expertise, and personality
- Spawn multiple instances from the same type (e.g., Dave, Ellis, Kate as Software Developers)
- Agent-to-agent communication and collaboration
- Per-instance customization with domain-specific knowledge
Complete SDLC Support
- Requirements Discovery → Architecture Spike → Implementation Planning
- Sprint Management with automated ceremonies
- Issue Tracking with Epic/Story/Task hierarchy
- Git Integration with proper branch/PR workflows
- CI/CD Pipelines with automated testing
Configurable Autonomy
- From
FULL_CONTROL(approve everything) toAUTONOMOUS(only major milestones) - Client can intervene at any point
- Transparent progress visibility
MCP-First Architecture
- All integrations via Model Context Protocol (MCP) servers
- Unified Knowledge Base with project/agent scoping
- Git providers (Gitea, GitHub, GitLab) via MCP
- Extensible through custom MCP tools
Project Complexity Wizard
- Script → Minimal process, no repo needed
- Simple → Single sprint, basic backlog
- Medium/Complex → Full AGILE workflow with multiple sprints
Technology Stack
Built on PragmaStack:
| Component | Technology |
|---|---|
| Backend | FastAPI 0.115+ (Python 3.11+) |
| Frontend | Next.js 16 (React 19) |
| Database | PostgreSQL 15+ with pgvector |
| ORM | SQLAlchemy 2.0 |
| State Management | Zustand + TanStack Query |
| UI | shadcn/ui + Tailwind 4 |
| Auth | JWT dual-token + OAuth 2.0 |
| Testing | pytest + Jest + Playwright |
Syndarix Extensions
| Component | Technology |
|---|---|
| Task Queue | Celery + Redis |
| Real-time | FastAPI WebSocket / SSE |
| Vector DB | pgvector (PostgreSQL extension) |
| MCP SDK | Anthropic MCP SDK |
Project Status
Phase: Architecture & Planning
See docs/requirements/ for the comprehensive requirements document.
Current Milestones
- Fork PragmaStack as foundation
- Create requirements document
- Execute architecture spikes
- Create ADRs for key decisions
- Begin MVP implementation
Documentation
- Requirements Document
- Architecture Decisions (coming soon)
- Spike Research (coming soon)
- Architecture Overview (coming soon)
Getting Started
Prerequisites
- Docker & Docker Compose
- Node.js 20+
- Python 3.11+
- PostgreSQL 15+ (or use Docker)
Quick Start
# Clone the repository
git clone https://gitea.pragmazest.com/cardosofelipe/syndarix.git
cd syndarix
# Copy environment template
cp .env.template .env
# Start development environment
docker-compose -f docker-compose.dev.yml up -d
# Run database migrations
make migrate
# Start the development servers
make dev
Architecture Overview
+====================================================================+
| SYNDARIX CORE |
+====================================================================+
| +------------------+ +------------------+ +------------------+ |
| | Agent Orchestrator| | Project Manager | | Workflow Engine | |
| +------------------+ +------------------+ +------------------+ |
+====================================================================+
|
v
+====================================================================+
| MCP ORCHESTRATION LAYER |
| All integrations via unified MCP servers with project scoping |
+====================================================================+
|
+------------------------+------------------------+
| | |
+----v----+ +----v----+ +----v----+ +----v----+ +----v----+
| LLM | | Git | |Knowledge| | File | | Code |
| Providers| | MCP | |Base MCP | |Sys. MCP | |Analysis |
+---------+ +---------+ +---------+ +---------+ +---------+
Contributing
See CONTRIBUTING.md for guidelines.
License
MIT License - see LICENSE for details.
Acknowledgments
- Built on PragmaStack
- Powered by Claude and the Anthropic API