Reading
A skill-based workflow tool designed for AI programming assistants such as Claude Code, Codex, and Gemini CLI. It enforces design, planning, TDD, and code reviews through an auto-triggered skill system, enabling AI assistants to behave like well-trained engineers.
Section 01
Arcforge is a skill-based workflow engine for AI programming assistants like Claude Code, Codex, and Gemini CLI, aiming to solve the problem of AI assistants' lack of discipline. Through an auto-triggered skill system, it embeds standardized software engineering practices (such as design, planning, TDD, and code reviews) into AI assistants' conversations, guiding them to work like well-trained engineers and improving code quality and maintainability.
Section 02
AI programming assistants have seen rapid improvement in capabilities, but they share a common issue: lack of discipline. When directly asked to implement features, AI often skips design, ignores reviews, and loses context, leading to code that works but accumulates technical debt quickly (e.g., design flaws, improper boundary handling, insufficient test coverage). Thus, the GregoryHo/arcforge project was born—it is not a tool to replace AI assistants, but a workflow engine that embeds standardized engineering practices into conversations via a skill system.
Section 03
Traditional AI assistants rely on users to remember commands and best practices. Arcforge encapsulates these practices into skills that are auto-triggered based on context, guiding AI to complete design, planning, TDD, review, and other steps, structuring conversations and reusing knowledge.
When a conversation starts, Arcforge hooks inject available skills and activate different skills based on task nature (e.g., arc-brainstorming for vague requirements, execution skills for defined tasks).
Section 04
Covers the full process from design to delivery: arc-brainstorming (design exploration) → arc-refining (refine requirements into specifications) → arc-planning (break down tasks and build dependency graphs) → arc-coordinating (parallel Git worktrees for large projects) → arc-implementing (enforce TDD).
After task completion, two stages of review are required: specification compliance review (functional completeness, boundary handling, etc.) and code quality review (style, design patterns, performance, etc.).
arc-journaling records conversation work, issues, and experiences; arc-reflecting analyzes logs to extract reusable patterns and form instincts.
Section 05
For vague requirements, first activate arc-brainstorming to clarify requirements and produce design documents; after approval, break down tasks.
arc-planning builds task dependency graphs, and arc-coordinating creates isolated Git worktrees to support parallel development.
arc-debugging initiates a systematic debugging process, and arc-tdd ensures fixes include regression tests.
arc-reflecting analyzes logs to extract experience patterns (e.g., time zone handling) and applies them to future conversations.
Section 06
Arcforge supports multiple platforms: Claude Code can install the plugin via marketplace sources; Codex and Gemini CLI require manual configuration; OpenCode requires cloning the repository and creating symbolic links.
Section 07
Arcforge pushes AI programming assistants from an ad-hoc Q&A mode to a structured engineering workflow. It does not replace creativity but ensures high-quality, maintainable code through discipline. For teams using Claude Code, Codex, or Gemini CLI, it provides a way to integrate AI into standardized processes. In the future, such tools will become key infrastructure to improve AI collaboration efficiency.
Keep going with more reads from the same topic.
Splinter is a minimalist, high-performance key-value (KV) and vector storage system enabling zero-latency inter-process communication via shared memory and atomic operations. With only 766 lines of core code, it supports millions of operations per second and 768-dimensional vector storage, offering a new architectural approach for local LLM inference and data-intensive applications.
Folkering OS is the world's first AI-native bare-metal operating system, entirely written in Rust no_std without relying on Linux, POSIX, or libc. It can generate commands from scratch, compile them into WASM, and run them in 10 seconds, achieving true self-evolution.
This project explores how to use Large Language Models (LLMs) to detect logical vulnerabilities in Ethereum smart contracts, especially business logic flaws that are difficult for traditional static analysis tools to capture, such as economic manipulation and execution order errors.
This article provides an in-depth analysis of the awesome-image-generation project—an authoritative list maintained by Backblaze Labs. It comprehensively covers commercial services, open-source weight models, development frameworks, and deployment infrastructure in the AI image generation domain, offering a systematic reference for developers building visual applications.