🏆 Winner of the 2025 TRAE Global Best Practice Challenge
Author: Marco — Full-Stack Developer & AI Workflow Architect
This repository is no longer maintained.
Due to a loss of trust in TRAE's management and their lack of transparency regarding platform decisions, I have decided to discontinue my collaboration with TRAE. As a result:
- ❌ No further updates will be made to this repository
- ❌ Issues will not be addressed - the repository is effectively archived
- ❌ No new content will be added
- 📂 This repository remains available as a reference for the community
- 🔄 Feel free to fork and continue the work independently
I believe in transparency and trust between platform providers and their community. When those values are compromised, it's time to move on.
- Main Whitepaper - Core concepts and architecture
- 10 Best Tips - Quick practical tips for immediate implementation
- SOLO + GLM-5 Best Combo - Deep dive into cost optimization
- Best Models Guide - Which model to choose for each task
- Billing Updates - New token-based membership details
- New Billing Stats - Real-world usage data and technical analysis
- Billing & Token Optimization - How to save tokens and reduce costs
- Rulesets Template - Ready-to-use ruleset examples
This repository documents the full engineering workflow I used to build, automate and scale software development tasks through TRAE, combined with custom-built AI Agents, strict Rule Systems, and an optimized model selection strategy based on performance benchmarks.
The core of this work is detailed in our Main Whitepaper, which explains how TRAE becomes a real AI Engineering Team, how to orchestrate multi-agent execution, and how leveraging GLM-5 (z.ai) inside TRAE dramatically reduces cost while increasing output efficiency.
- 💡 Best Models for TRAE - A complete comparison of models (Gemini, GPT, Kimi) and which one to choose for frontend, backend, or refactoring.
- ⚡ SOLO + GLM-5: Best Combo - Comprehensive deep-dive on the most cost-effective TRAE setup using GLM-5 to save up to 100x on credits.
- 📋 Rulesets Template - Ready-to-use rule templates and system prompts for deterministic agent behavior.
- 💡 10 Best Tips - Quick, actionable tips for immediate implementation and workflow optimization.
- 📉 Billing & Token Optimization - Detailed strategies for saving tokens, reducing costs, and maximizing your requests.
- 💰 Billing & Membership Updates - Everything you need to know about the transition from request-based to token-based usage.
- 📈 New Billing Stats Report - Real-world technical data, model efficiency, and token burn analysis from intensive usage.
- 🤖 Model Availability - Why certain models are restricted and what are the best alternatives.
- 🐧 Linux Version Status - Latest info on the upcoming (and currently in internal testing) Linux release.
- Agents VS Rules VS Skills - Understand the core differences in TRAE.
- TRAE-Agents Collection - Production-ready AI agents for the TRAE ecosystem.
- TRAE Skills Repository - Tips and tricks for TRAE.
- Get 10% Discount on GLM-5 - Use this referral link for 10% off the Coding Plan.
This project is licensed under the MIT License - see the LICENSE file for details.
| Metric | Status |
|---|---|
| Maintenance | ⛔ Discontinued |
| Issues | ❌ Not monitored |
| Pull Requests | ❌ Not accepted |
| Content Updates | ❌ None planned |
| Repository | 📂 Available as reference |
| Forks | ✅ Allowed and encouraged |
Thank you to everyone who supported, starred, and contributed to this repository. The TRAE community is amazing, and the knowledge shared here will remain available for anyone who finds it valuable.
For those continuing with TRAE, I wish you the best in your journey. Always prioritize platforms that value transparency and trust their community.
Last Updated: 2026
This repository is now archived. Fork it to continue the work.