#AI编程
9513 related articles
game-superpowers: A Game Development Skill Library for AI Coding Tools
game-superpowers is an open-source toolkit that adds game development skills to AI coding tools like Claude Code, Codex, and Gemini CLI via reusable workflows for auditing, polishing, and repairing game projects.
Agent-Dev-Team: A Deep Dive into the Tool-Agnostic Hierarchical Engineering Agent Framework
agent-dev-team is a tool-agnostic, hierarchical multi-agent framework with 21 engineering roles, seniority ceilings, HANDOFF packets, and portable skills — works alongside LangChain.
The Multi-Agent Architecture Debate: Fixed Graph vs. Dynamic Discovery
Fixed graph vs. dynamic discovery in multi-agent systems? This article breaks down both architectures, covers six key challenges of dynamic delegation, and offers practical guidance for LangGraph/LangChain teams.
AI Coding Model Benchmark Tool: GPT-5.3 Codex vs. Claude Opus 4.6 — Which One Wins?
The open-source project ai-coding-benchmark-zyt benchmarks GPT-5.3 Codex vs. Claude Opus 4.6. This article explores its methodology, value, and developer guidance.
Three AI Coding Models Head-to-Head: Opus 4.7, GPT-5, and DeepSeek V4-Pro Benchmarked
An open-source GitHub project pits Opus 4.7, GPT-5, and DeepSeek V4-Pro against each other building a threshold ECDSA CLI tool in Rust, revealing real-world AI coding performance.
GPT-5 Codex vs Claude Sonnet 4: A Code Quality Experiment for AI Programming
An open-source experiment on GitHub compares GPT-5 Codex vs Claude Sonnet 4 on Java code quality, exploring evaluation methods and model selection for AI coding assistants.
Tired of Complex Workflows? A Roundup of Platforms for Building AI Agents with Natural Language
Find n8n and Make too complex? This article surveys natural language AI Agent builder platforms, breaks down two key approaches, and offers practical selection tips for non-technical users.
Real-SWE: Benchmarking AI Coding Models on Real Enterprise Codebases
Real-SWE benchmarks AI coding models on private enterprise codebases to avoid open-source data contamination, offering a more realistic measure of model performance on unfamiliar code.
Ludwig: Building Custom LLMs and Neural Networks with Declarative Low-Code Configuration
Ludwig is an open-source low-code ML framework that builds custom LLMs and neural networks via declarative YAML config — no boilerplate training code required.
Cline: A Deep Dive into the Autonomous AI Coding Agent Inside Your IDE
Cline is an open-source autonomous coding agent in your IDE that creates files, runs commands, and controls browsers — with developer approval at every step.
From Figma to React: Observations on AI-Generated Admin Dashboards
DashStack Admin Dashboard on GitHub was auto-converted from a Figma design to React + TypeScript using Claude Code. A look at what these AI "design-to-code" eval projects reveal.
CadQuery vs. OpenSCAD: Which Is Better for AI Agentic CAD Work?
CadQuery vs. OpenSCAD for AI agentic CAD: comparing Python ecosystem, error feedback, and geometric expressiveness to find the best tool for AI-driven 3D modeling.
Void Linux Maintainer Orphans ~100 Packages Over AI Policy Dispute
A Void Linux maintainer orphaned ~100 packages over an AI policy dispute, highlighting open source governance challenges around AI contributions, licensing, and community values.
Oumi: An All-in-One Open-Source Framework for LLM Fine-Tuning, Evaluation, and Deployment
Oumi is an open-source Python framework for the full LLM/VLM lifecycle — fine-tuning, evaluation, and deployment — with support for Gemma, Qwen, DeepSeek-R1, and more.
pi: A One-Stop AI Agent Toolkit — Architecture Breakdown Behind 46k GitHub Stars
pi (earendil-works/pi) is an open-source one-stop AI Agent toolkit integrating a coding agent CLI, unified LLM API, TUI & web UI libraries, Slack bot, and vLLM inference pods — with 46k+ GitHub Stars.
Long-Running Agent Framework: Enabling AI Agents to Handle Complex, Extended Tasks
long-running-agent-framework is an open-source JavaScript orchestration framework for AI agents to handle complex, long-horizon tasks with state persistence and reliable scheduling.
Goose: The Open-Source AI Agent That Goes Beyond Code Suggestions to Execute, Edit, and Test
Goose is an open-source, extensible AI Agent written in Rust that goes beyond code suggestions to autonomously install, execute, edit, and test code. Compatible with any LLM. 44K+ GitHub Stars.
dev-crew: An Out-of-the-Box AI Agent Software Development Team Orchestration Framework
dev-crew is an open-source AI Agent software development team orchestration framework. Uses CLI and Skill modules to enable multi-agent collaborative development with TypeScript.
Turbo-Flow: An Intelligent Development Environment Integrating 600+ AI Sub-Agents
turbo-flow is an open-source agentic dev environment integrating 600+ AI subagents, Claude Flow, and SPARC methodology with support for GitHub Codespaces, Google Cloud Shell, and Rackspace Spot.
Jules AI Autonomous Agent Prompt Library: Let AI Optimize Your Codebase Around the Clock
The open-source project Jules-Google-AI-autonomous-agents offers copy-paste prompts for deploying autonomous AI agents to continuously optimize, harden, and enhance your codebase.