Summary

HelloAGENTS is a modular AI programming skill system that transforms chaotic AI agent outputs into structured, traceable, production-ready code. It targets teams and developers needing systematic requirement validation, transparent AI decision-making, and robust cross-platform compatibility, offering intelligent routing and human-centric workflows to adapt dynamically to task complexity.

How It Works

The system employs multi-dimensional analysis—semantic understanding, intent classification, scope estimation, and risk detection—to intelligently route user requests to one of four specialized workflows. A core differentiator is its 10-point requirement scoring system, which enforces clarity and completeness before development begins, prompting targeted follow-up questions when scores fall below a threshold. This is complemented by a 3-phase workflow (Analysis, Design, Development) and G12 state variables that maintain context across interactions, ensuring systematic progression and traceability. Human-centric safeguards, including explicit uncertainty disclosure (G3) and Extreme High-Risk Behavior (EHRB) escalation, enhance safety and transparency.

Quick Start & Requirements

Installation involves copying the ruleset to a specified directory within your CLI environment (e.g., ~/.codex/Skills/EN for Codex CLI). Users must configure the OUTPUT_LANGUAGE in the AGENTS.md header to "English" for consistent operation. Prerequisites include a CLI environment with file system access.

Highlighted Details

• Unified Intelligent Routing: Automatically selects one of four workflows (Quick Fix, Light Iteration, Standard Development, Full R&D) based on semantic analysis, intent, scope, and risk detection.
• Requirements Analysis with Scoring: A 10-point system across four dimensions (Goal Clarity, Expected Results, Scope Boundaries, Constraints) ensures requirement quality, triggering targeted follow-up questions for scores below 7.
• Phase & State Management: A systematic 3-phase workflow (Analysis → Design → Development) uses G12 state variables to preserve context across interactions, ensuring traceability and consistent execution.
• Human-Centric Safeguards: Features G3 Uncertainty Principles for transparent decision-making, EHRB escalation for safety, and a unified, validated output format.
• Cross-Platform Compatibility: Supports Windows PowerShell, macOS, and Linux environments.

Maintenance & Community

The project actively encourages contributions via pull requests and provides channels for bug reports and feature requests through GitHub Issues and Discussions, respectively. Project stats indicate active development and community interest.

Licensing & Compatibility

Code is licensed under Apache License 2.0, and documentation under CC BY 4.0. Both allow commercial use, provided attribution requirements are met. Apache 2.0 requires retaining LICENSE and NOTICE, while CC BY 4.0 necessitates indicating changes and providing a license link.

Limitations & Caveats

The system is not intended for one-off scripts lacking quality requirements or environments without file system access. Proper configuration, including setting the output language, is crucial for expected behavior. Ambiguous requests or incomplete requirement dimensions can lead to incorrect workflow routing or persistent follow-up questions.