New conversation

Philippe Laurent

Tech serving business, never the other way around

Ask anything about this project, the agents will answer...

Vision

Software Factory — Real Agentic ≠ Workflow Automation.
Agents collaborent (débat/véto/délégation) pour produire du code.
128 agents SAFe, 23 patterns, 12 workflows, 1222 skills.
Stack: FastAPI + Jinja2 + HTMX + SSE + SQLite (WAL + FTS5).
Dark purple theme. Zero emoji. SVG Feather icons.

2 Epics:
1. NEW FEATURES — Discovery → Comité Strat → Arch → Dev Sprint → QA → Deploy → Retro
2. TMA — Détection → Triage P0-P4 → Diagnostic → Fix TDD → Non-Régression → Hotfix Deploy

Values

quality feedback no-waste kaizen flow tdd

Lead Agent

Philippe Laurent

DSI / CTO

Domains

backend frontend agents patterns infra

Memory

CLAUDE.md 16.9K

MACARON AGENT PLATFORM

WHAT

Web multi-agent platform SAFe-aligned. Agents collaborate (debate/veto/delegate) autonomously.
FastAPI + HTMX + SSE + SQLite. Dark purple theme. Port 8099.

RUN

cd _SOFTWARE_FACTORY
python3 -m uvicorn platform.server:app --host 0.0.0.0 --port 8099 --ws none --log-level warning
# NO --reload (conflicts stdlib `platform` module)
# --ws none mandatory (SSE not WS)
# DB: data/platform.db (parent dir, NOT platform/data/)
# ⚠️ NEVER rm -f data/platform.db — persistent, contains missions/sessions/memory

⚠️ CRITICAL RULES

NEVER delete data/platform.db — init_db() handles migrations idempotently
NEVER set *_API_KEY=dummy — real keys loaded from ~/.config/factory/*.key
NEVER import platform at top-level (shadows stdlib)
**N

… (truncated)

.github/copilot-instructions.md 6.0K

Software Factory — Copilot Instructions

Build & Run

# Install dependencies
pip install -r requirements.txt

# Run locally (from the parent _SOFTWARE_FACTORY/ directory)
python -m uvicorn platform.server:app --host 0.0.0.0 --port 8090 --ws none

# IMPORTANT: Do NOT use --reload (conflicts with stdlib `platform` module name)
# IMPORTANT: --ws none is required (SSE is used instead of WebSockets)
# The DB is auto-created at startup in platform/data/platform.db

Required env vars:

AZURE_OPENAI_API_KEY      # Azure OpenAI (primary LLM)
AZURE_OPENAI_ENDPOINT     # e.g. https://castudioiatestopenai.openai.azure.com/
# Optional fallback providers:
MINIMAX_API_KEY           # MiniMax M2.5
GLM_API_KEY               # Zhipu GLM-4.7

Architecture

**FastAPI + HTMX + SS

… (truncated)

SPECS.md 35.1K

MACARON AGENT PLATFORM — SPECIFICATIONS v1.0

Vision : Une plateforme web locale permettant de créer, orchestrer et piloter des équipes d'agents IA collaboratifs sur des projets logiciels, avec mémoire persistante, outils MCP, et conversation temps réel style WhatsApp.

1. CONCEPTS FONDAMENTAUX

1.1 Projet = Unité de base

Tout commence par un Projet. Un projet est un espace de travail complet :

```
PROJET
├── Identity → nom, description, avatar, couleur
├── Vision → document de vision produit (VISION.md)
├── Values → principes Lean/Agile du projet (valeurs)
├── Git → répertoire local, historique, branches
├── Agent Lead → agent LLM par défaut (interlocuteur humain)
├── Agents → pool d'agents disponibles pour ce projet
├── Patterns

… (truncated)

README.md 7.5K

🍪 Software Factory

Multi-agent collaborative platform web — agents spécialisés (Métier, Lead Dev, Testeur, Sécurité, DevOps...) qui dialoguent entre eux via le protocole A2A, orchestrés selon 8 patterns agentic, hébergé sur VM Azure avec Azure AI Foundry.

Quick Start

# Install deps
cd platform
pip install -r requirements.txt

# Run locally
python -m uvicorn platform.server:app --port 8090 --reload

# Open http://localhost:8090

Architecture

```
┌─── Web UI (HTMX + SSE) ────────────────────────────────┐
│ Workspace │ Agent Chat │ Conversation │ Skills │ Monitor │
└─────────────────────────────────────────────────────────┘
│ SSE (real-time) │ REST API
┌────────┴──────────────────────────┴────────────────────┐
│ ORCHESTRATOR (Python)

… (truncated)

Loading git...

Loading tasks...