🎉 QonQrete v1.0.0-stable Released!

December 30, 2025 | Major Release Announcement

After months of intensive development, testing, and 90 validation runs across our WoNQ Matrix testing framework, we’re thrilled to announce the production-ready release of QonQrete v1.0.0-stable!

What is QonQrete?

QonQrete is an open-source, local-first multi-agent AI orchestration system that generates code autonomously through iterative cycles. Think of it as a self-hosted alternative to cloud-based AI coding platforms, giving you complete control over your AI-powered development workflow.

The Three-Agent Pipeline

📝 InstruQtor    →    🔨 ConstruQtor    →    🔍 InspeQtor
   (Planning)            (Building)           (Reviewing)

InstruQtor breaks down your task into manageable “briqs” (building blocks)
ConstruQtor generates the actual code for each briq
InspeQtor reviews the output and identifies issues
Repeat until your project is complete!

What’s New in v1.0.0-stable?

🔥 Bulletproof Language Detection

The #1 issue reported during beta was the “py file bug” - when using certain AI providers (especially OpenAI), code blocks tagged with shorthand identifiers like ```py would create files literally named “py” instead of being recognized as Python code.

The fix: We expanded our language detection from 23 keywords to 400+ entries, covering:

Every variant of every major programming language
All GitHub Linguist identifiers
Infrastructure-as-Code languages (Terraform, Ansible, Kubernetes, etc.)
Generic markers AI models use (code, snippet, output, console)

Result: QonQrete now works flawlessly with ALL major AI providers:

Provider	Status
✅ OpenAI GPT-4/GPT-4o	NOW BULLETPROOF
✅ Google Gemini	Works perfectly
✅ Anthropic Claude	Works perfectly
✅ DeepSeek	NOW BULLETPROOF
✅ Qwen	NOW BULLETPROOF

🎯 Enforced Briq Sensitivity

Previously, the sensitivity setting was just a “hint” to the AI. Now it’s enforced:

Sensitivity	Briq Range	Use Case
9	Exactly 1	Single-file scripts
7	3-5 briqs	Recommended default
5	8-12 briqs	Component-level work
3	15-20 briqs	Standard granularity
0	40-60 briqs	Maximum decomposition

If the AI produces too few briqs, the system retries with a stronger prompt. Too many? It intelligently merges them. Consistency guaranteed.

🆕 New CLI Controls

# Override cycle count directly
./qonqrete.sh run -c 6

# Combine with sensitivity
./qonqrete.sh run -b 5 -c 6

Battle-Tested: The WoNQ Matrix

We didn’t just build it and ship it. We validated it.

The WoNQ Matrix is our comprehensive testing framework that runs QonQrete across every combination of sensitivity levels (0-9) and cycle counts (1-9). That’s 90 unique configurations, each tested with a complex real-world project.

Results

╔═══════════════════════════════════════════════════════════════╗
║                 WoNQ MATRIX RESULTS                           ║
╠═══════════════════════════════════════════════════════════════╣
║ Total Runs:           90 (100% coverage)                      ║
║ Clean Completions:    90 (100% success)                       ║
║ Champion Score:       658 (sensitivity=3, cycle=7)            ║
║ Global Average:       554                                     ║
║ Scores ≥600:          35.6%                                   ║
╚═══════════════════════════════════════════════════════════════╝

Key Findings

Sweet Spot: Sensitivity 2-4 with 5-7 cycles produces the best results
Champion Config: sens=3, cycle=7 consistently hits peak performance
Avoid: Sensitivity ≥7 shows diminishing returns; Cycle 8 is a “death valley”

Cost Efficiency

QonQrete is designed to be cost-efficient. With our batched review system and local-first architecture:

Configuration	Estimated Cost per Run
Simple project (sens=7, 4 cycles)	~$0.50
Medium project (sens=5, 6 cycles)	~$2.00
Complex project (sens=3, 7 cycles)	~$4.00

Compare that to manually prompting an AI assistant dozens of times!

Getting Started

Prerequisites

Docker
An API key from any supported provider (OpenAI, Gemini, Anthropic, DeepSeek, or Qwen)

Quick Start

# Clone the repo
git clone https://github.com/qonqrete/qonqrete.git
cd qonqrete

# Copy env template and add your API keys
cp .env.example .env
nano .env

# Initialize (builds Docker container)
./qonqrete.sh init

# Create your task
echo "Build a REST API with FastAPI..." > worqspace/tasq.md

# Run!
./qonqrete.sh run

Recommended Configuration

For most projects, we recommend:

# config.yaml
briq_sensitivity: 7
auto_cycle_limit: 4

For complex multi-service projects:

briq_sensitivity: 5
auto_cycle_limit: 6

What’s Next?

v1.0.0-stable is a major milestone, but we’re not stopping here! On the roadmap:

🔌 MCP Server Integration - Full Model Context Protocol support
🧠 Enhanced Context Management - Smarter caching and context window optimization
🔄 Parallel Briq Execution - Speed up large projects with concurrent building
📊 Built-in Analytics Dashboard - Visualize your build metrics

🔗 Links

GitHub: github.com/illdynamics/qonqrete
Docs: qonqrete.sh/docs
Discord: Coming soon!
Twitter/X: @QonQrete

Found a bug? Have a feature request? Open an issue on GitHub!

Thank You!

A massive thank you to everyone who tested the beta versions, reported bugs, and provided feedback. This release wouldn’t be possible without you.

Now go build something amazing with QonQrete v1.0.0-stable! 🚀

Changelog Highlights (v0.9.9-beta → v1.0.0-stable)

+400 language identifiers for bulletproof code block parsing
Enforced briq sensitivity with automatic retry/merge logic
New -c/--cyqles CLI flag for cycle override
Updated defaults: sensitivity 7, cycles 4
90 validation runs confirming production readiness

QonQrete is open-source software released under the MIT License.

#QonQrete #LocalFirst #AIAgents #OpenSource #v1.0.0

🎉 QonQrete v1.0.0-stable Released! The Production-Ready AI Orchestration Platform