AI Agent Directory

Zero-Config Code Flow for Claude code & Codex

The agent that grows with you

《御舆：解码 Agent Harness》42万字拆解 AI Agent 的Harness骨架与神经 —— Claude Code 架构深度剖析，15 章从对话循环到构建你自己的 Agent Harness。在线阅读网站：

"🐈 nanobot: The Ultra-Lightweight Personal AI Agent"

AI data analyst. Upload CSV, Excel, SQL databases — get visualizations, insights, and statistical analysis in plain English. No coding required.

awesome-agents — discovered from curated AI agents list.

https://adongwanai.github.io/AgentGuide | AI Agent开发指南 | LangGraph实战 | 高级RAG | 转行大模型 | 大模型面试 | 算法工程师 | 面试题库 | 强化学习｜数据合成

The seminal task-driven autonomous agent. 140 lines of Python that started the agentic AI revolution. Uses GPT-4 and vector memory.

Convert NotebookLM PDFs to PPTX with separated background images and editable text layers using Gemini AI

The SOTA Open-Source Browser Agent for autonomously performing complex tasks on the web

ARIS ⚔️ (Auto-Research-In-Sleep) — Lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, and experiment auto...

Stop prompting. Start specifying.

A wonderful list of Game Development resources.

SmolLM2 135M — a text-generation model by undefined on HuggingFace. 1,311,189 downloads, 179 likes.

Qwen3 Coder 30B A3B Instruct — a text-generation model by undefined on HuggingFace. 1,310,727 downloads, 1004 likes.

DeepSeek R1 Distill Llama 8B — a text-generation model by undefined on HuggingFace. 1,753,480 downloads, 844 likes.

Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token ...

Llama 2 7b Hf — a text-generation model by undefined on HuggingFace. 1,373,140 downloads, 2290 likes.

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window...

Elephant Alpha is a 100B-parameter text model focused on intelligence efficiency, delivering strong reasoning performance while minimizing token usage. It supports a 256K context window with up to 32K output...