wehub/lsdefine--genericagent
1
0
Fork 0
Code Issues Pull Requests Actions Deployments Models Agent Notes Packages Projects Releases Wiki Activity

docs: make Chinese README the default

Browse Source
This commit is contained in:
wehub-resource-sync
2026-07-13 10:13:04 +00:00
parent e30fa57e89
commit 7dbb68934e
1 changed files with 162 additions and 163 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+162 -163
View File
@@ -1,12 +1,18 @@
<!-- WEHUB_ZH_README -->
> [!NOTE]
> 本文档由 WeHub 基于上游 README 翻译整理，属于社区翻译，非官方中文文档。
> [English](./README.en.md) · [原始项目](https://github.com/lsdefine/GenericAgent) · [上游 README](https://github.com/lsdefine/GenericAgent/blob/HEAD/README.md)
> 原作者、版权与许可证归属以原始项目及本仓库 LICENSE 文件为准。
<div align="center">
<img src="assets/images/bar.jpg" width="880" alt="GenericAgent Banner"/>
# GenericAgent
**A Minimal, Self-Evolving Autonomous Agent Framework**
**极简、可自我进化的自主智能体框架**
*~3K lines of seed code · 9 atomic tools · ~100-line Agent Loop*
*约 3K 行种子代码 · 9 个原子工具 · 约 100 行的 Agent Loop*
<p>
@@ -25,111 +31,111 @@
</div>
> 📌 **Official:** GitHub + https://gaagent.ai only. DintalClaw is the sole authorized commercial partner; others are not affiliated.
> 📌 **官方渠道：** 仅 GitHub + https://gaagent.ai。DintalClaw 是唯一授权商业合作伙伴；其他均非关联方。
---
<a id="-english"></a>
## 🌟 Overview
## 🌟 概览
**GenericAgent** is a minimal, self-evolving autonomous agent framework. Its core is just **~3K lines of code**. Through **9 atomic tools + a ~100-line Agent Loop**, it grants any LLM system-level control over a local computer — covering browser, terminal, filesystem, keyboard/mouse input, screen vision, and mobile devices (ADB).
**GenericAgent** 是一个极简、可自我进化的自主智能体（Autonomous Agent）框架。其核心仅 **约 3K 行代码**。通过 **9 个原子工具 + 约 100 行的 Agent Loop**，它赋予任意 LLM 对本地计算机的系统级控制能力——覆盖浏览器、终端、文件系统、键盘/鼠标输入、屏幕视觉以及移动设备（ADB）。
> Design philosophy — **don't preload skills, evolve them.**
> 设计理念 —— **不预加载技能，而是让它们进化。**
Every time GenericAgent solves a new task, it automatically crystallizes the execution path into a reusable **Skill**. The longer you use it, the more skills accumulate — forming a personal skill tree grown entirely from 3K lines of seed code.
每当 GenericAgent 完成一项新任务，它会自动将执行路径凝结为可复用的 **Skill（技能）**。使用越久，积累的技能越多——最终形成一棵完全由 3K 行种子代码生长而来的个人技能树。
> 🤖 **Self-Bootstrap Proof** — Everything in this repository, from installing Git and running `git init` to every commit message, was completed autonomously by GenericAgent. The author never opened a terminal once.
> 🤖 **自举证明（Self-Bootstrap Proof）** —— 本仓库中的一切，从安装 Git、运行 `git init`，到每一条 commit message，均由 GenericAgent 自主完成。作者从未打开过终端。
### 📑 Table of Contents
### 📑 目录
- [Key Features](#-key-features)
- [Demo Showcase](#-demo-showcase)
- [Quick Start](#-quick-start)
- [Usage](#-usage)
- [Unlocking Advanced Capabilities](#-unlocking-advanced-capabilities)
- [Architecture](#-architecture)
- [Self-Evolution Mechanism](#-self-evolution-mechanism)
- [Comparison](#-comparison)
- [Evaluation](#-evaluation)
- [Roadmap & News](#-roadmap--news)
- [Community & Support](#-community--support)
- [License](#-license)
- [核心特性](#-key-features)
- [演示展示](#-demo-showcase)
- [快速开始](#-quick-start)
- [使用方式](#-usage)
- [解锁高级能力](#-unlocking-advanced-capabilities)
- [架构](#-architecture)
- [自我进化机制](#-self-evolution-mechanism)
- [对比](#-comparison)
- [评测](#-evaluation)
- [路线图与动态](#-roadmap--news)
- [社区与支持](#-community--support)
- [许可证](#-license)
---
## 📋 Key Features
## 📋 核心特性
| Feature | Description |
| 特性 | 说明 |
| :--- | :--- |
| 🧬 **Self-Evolving** | Automatically crystallizes each task into a Skill. Capabilities grow with every use, forming your personal skill tree. |
| 🪶 **Minimal Architecture** | ~3K lines of core code. Agent Loop is ~100 lines. No complex dependencies, zero deployment overhead. |
| ⚡ **Strong Execution** | **TMWebdriver** injects into a real browser (preserving login sessions). 9 atomic tools take direct control of the system. |
| 🔌 **High Compatibility** | Supports Claude / Gemini / Kimi / MiniMax and other major models. Cross-platform. |
| 💰 **Token Efficient** | <30K context window — a fraction of the 200K–1M other agents consume. Less noise, fewer hallucinations, higher success rate, lower cost. |
| 🧬 **自我进化** | 自动将每项任务凝结为 Skill。能力随使用不断增长，形成你的个人技能树。 |
| 🪶 **极简架构** | 核心代码约 3K 行。Agent Loop 约 100 行。无复杂依赖，零部署开销。 |
| ⚡ **强执行力** | **TMWebdriver** 注入真实浏览器（保留登录会话）。9 个原子工具直接控制系统。 |
| 🔌 **高兼容性** | 支持 Claude / Gemini / Kimi / MiniMax 及其他主流模型。跨平台。 |
| 💰 **Token 高效** | 上下文窗口 <30K —— 仅为其他智能体消耗的 200K–1M 的一小部分。更少噪声、更少幻觉、更高成功率、更低成本。 |
---
## 🎯 Demo Showcase
## 🎯 演示展示
<table>
<tr>
<td align="center" width="50%"><b>🛡️ Real-Browser CAPTCHA Survival</b></td>
<td align="center" width="50%"><b>🌐 Autonomous Web Exploration</b></td>
<td align="center" width="50%"><b>🛡️ 真实浏览器 CAPTCHA 通关</b></td>
<td align="center" width="50%"><b>🌐 自主网页探索</b></td>
</tr>
<tr>
<td><img src="assets/demo/discord_hcaptcha_real_browser.gif" width="100%" alt="Discord hCaptcha passed in real browser"></td>
<td><img src="assets/demo/autonomous_explore.png" width="100%" alt="Web Exploration"></td>
</tr>
<tr>
<td><sub>While configuring a Discord bot, an hCaptcha <i>"Are you human?"</i> challenge pops up mid-task — GA's real browser session passes it and the task continues. See <a href="#browser-realness-of-ga-web-tools">Browser Realness</a>.</sub></td>
<td><sub>Autonomously browses and periodically summarizes web content.</sub></td>
<td><sub>配置 Discord 机器人时，任务中途弹出 hCaptcha <i>"Are you human?"</i> 验证 —— GA 的真实浏览器会话通过验证，任务继续。参见 <a href="#browser-realness-of-ga-web-tools">Browser Realness</a>。</sub></td>
<td><sub>自主浏览并定期汇总网页内容。</sub></td>
</tr>
<tr>
<td align="center"><b>🧋 Food Delivery Order</b></td>
<td align="center"><b>📈 Quantitative Stock Screening</b></td>
<td align="center"><b>🧋 外卖点单</b></td>
<td align="center"><b>📈 量化选股</b></td>
</tr>
<tr>
<td><img src="assets/demo/order_tea.gif" width="100%" alt="Order Tea"></td>
<td><img src="assets/demo/selectstock.gif" width="100%" alt="Stock Selection"></td>
</tr>
<tr>
<td><sub><i>"Order me a milk tea"</i> — navigates the delivery app, selects items, completes checkout.</sub></td>
<td><sub><i>"Find GEM stocks with EXPMA golden cross, turnover &gt; 5%"</i> — quantitative screening.</sub></td>
<td><sub><i>"Order me a milk tea"</i> —— 导航外卖应用、选择商品、完成结账。</sub></td>
<td><sub><i>"Find GEM stocks with EXPMA golden cross, turnover &gt; 5%"</i> —— 量化筛选。</sub></td>
</tr>
<tr>
<td align="center"><b>💰 Expense Tracking</b></td>
<td align="center"><b>💬 Batch Messaging</b></td>
<td align="center"><b>💰 记账查账</b></td>
<td align="center"><b>💬 批量发消息</b></td>
</tr>
<tr>
<td><img src="assets/demo/alipay_expense.png" width="100%" alt="Alipay Expense"></td>
<td align="center"><img src="assets/demo/wechat_batch.png" width="65%" alt="WeChat Batch"></td>
</tr>
<tr>
<td><sub><i>"Find expenses over ¥2K in the last 3 months"</i> — drives Alipay via ADB.</sub></td>
<td><sub>Sends bulk WeChat messages, fully driving the WeChat client.</sub></td>
<td><sub><i>"Find expenses over ¥2K in the last 3 months"</i> —— 通过 ADB 驱动支付宝。</sub></td>
<td><sub>批量发送微信消息，完全驱动微信客户端。</sub></td>
</tr>
</table>
---
## 🚀 Quick Start
## 🚀 快速开始
> ⚠️ **Python version**: use **Python 3.11 or 3.12**. **Do not** use Python 3.14 — it is incompatible with `pywebview` and a few other GA dependencies.
> ⚠️ **Python 版本**：请使用 **Python 3.11 或 3.12**。**不要**使用 Python 3.14 —— 它与 `pywebview` 及少数其他 GA 依赖不兼容。
>
> 📖 Detailed installation guide: **[installation.md](docs/installation.md)** · **[installation_zh.md（中文）](docs/installation_zh.md)**
> 📖 详细安装指南：**[installation.md](docs/installation.md)** · **[installation_zh.md（中文）](docs/installation_zh.md)**
### For LLM Agents
### 面向 LLM 智能体
Fetch the installation guide and follow it:
获取安装指南并按步骤操作：
```bash
curl -fsSL https://raw.githubusercontent.com/lsdefine/GenericAgent/refs/heads/main/docs/installation.md
```
### For Humans
### 面向人类用户
#### Method 1 — Clone & install *(recommended)*
#### 方法 1 — 克隆并安装 *（推荐）*
```bash
git clone https://github.com/lsdefine/GenericAgent.git && cd GenericAgent
@@ -137,18 +143,18 @@ uv venv && uv pip install -e ".[ui]"
cp mykey_template_en.py mykey.py # fill in your LLM API key
```
Dependencies are deliberately tiered: the agent core needs only `requests`, plus four lightweight packages (`beautifulsoup4`, `bottle`, `simple-websocket-server`, `aiohttp`) for TMWebdriver's local server. The `[ui]` extra pulls in frontend libraries (Streamlit, `prompt_toolkit`/`rich` for the TUI, …) — install it for the bundled UIs, or skip it entirely and drive the agent headless. No Playwright, no LangChain, no browser binaries to download.
依赖刻意分层：智能体核心仅需 `requests`，另加四个轻量包（`beautifulsoup4`、`bottle`、`simple-websocket-server`、`aiohttp`）用于 TMWebdriver 本地服务。`[ui]` 额外依赖会拉取前端库（Streamlit、用于 TUI 的 `prompt_toolkit`/`rich`，……）—— 若需捆绑 UI 请安装，也可完全跳过并以无头模式驱动智能体。无需 Playwright、LangChain，也无需下载浏览器二进制文件。
Then launch:
然后启动：
```bash
python frontends/tui_v3.py # Terminal UI (recommended)
python launch.pyw # Streamlit web UI
```
#### Method 2 — One-line installer *(convenience)*
#### 方法 2 — 一行安装 *（便捷）*
Sets up a self-contained directory with an isolated Python environment, Git, and a ready-to-run package. The script is in [`assets/`](assets/) if you'd like to read it first.
会搭建一个自包含目录，内含隔离的 Python 环境、Git 和可直接运行的包。脚本位于 [`assets/`](assets/)，如需可先阅读。
**Windows PowerShell**
@@ -162,31 +168,31 @@ powershell -ExecutionPolicy Bypass -c "$env:GLOBAL=1; irm https://raw.githubuser
GLOBAL=1 bash -c "$(curl -fsSL https://raw.githubusercontent.com/lsdefine/GenericAgent/main/assets/ga_install.sh)"
```
> 💡 GenericAgent grows its environment **through the Agent itself** — don't pre-install everything. See [Unlocking Advanced Capabilities](#-unlocking-advanced-capabilities) below.
> 💡 GenericAgent 通过 **智能体自身** 来扩展环境 —— 不要预先安装一切。参见下方 [解锁高级能力](#-unlocking-advanced-capabilities)。
---
## 💻 Usage
## 💻 使用方式
### Frontends
### 前端
#### Terminal UI *(recommended)*
#### 终端 UI *（推荐）*
A lightweight, scrollback-first terminal interface built on `prompt_toolkit` + `rich`. Supports multiple concurrent sessions and real-time streaming.
基于 `prompt_toolkit` + `rich` 构建的轻量、以滚动回显为先的终端界面。支持多会话并发与实时流式输出。
```bash
python frontends/tui_v3.py
```
<details>
<summary><b>⚠️ Windows TUI Troubleshooting</b></summary>
<summary><b>⚠️ Windows TUI 故障排查</b></summary>
TUI rendering on Windows can be flaky depending on terminal + font. Common causes:
Windows 上的 TUI 渲染可能因终端与字体而异而不稳定。常见原因：
1. `prompt_toolkit` / `rich` are not on the latest version — `pip install -U prompt_toolkit rich` first.
2. PowerShell / cmd ship with terminals that have rough Unicode + key-binding support. **Prefer Git Bash on Windows**, which is much better behaved.
3. If it still looks broken, ask GA itself to fix it:
> *"My experience using `frontends/tui_v3.py` in PowerShell / cmd / Git Bash on Windows is very poor — lots of incompatibility. Please refer to Claude Code's best practices for the Windows terminal and fix all font and rendering incompatibilities."*
1. `prompt_toolkit` / `rich` 不是最新版本 — 请先使用 `pip install -U prompt_toolkit rich`。
2. PowerShell / cmd 自带的终端对 Unicode 与快捷键支持较差。**在 Windows 上更推荐使用 Git Bash**，表现要好得多。
3. 如果显示仍然有问题，可以让 GA 自行修复：
> *"我在 Windows 的 PowerShell / cmd / Git Bash 中使用 `frontends/tui_v3.py` 的体验非常差 — 存在大量不兼容。请参考 Claude Code 针对 Windows 终端的最佳实践，修复所有字体与渲染不兼容问题。"*
</details>
@@ -198,7 +204,7 @@ python launch.pyw
### Bot Interface (IM)
GenericAgent also supports IM frontends such as Telegram, Discord, and Lark.
GenericAgent 还支持 Telegram、Discord、Lark 等 IM 前端。
| Platform | Command |
| :--- | :--- |
@@ -206,90 +212,83 @@ GenericAgent also supports IM frontends such as Telegram, Discord, and Lark.
| Discord | `python frontends/dcapp.py` |
| Lark / Feishu | `python frontends/fsapp.py` |
> WeChat, QQ, WeCom and DingTalk are also supported — see the Chinese section below.
> For detailed setup, ask GenericAgent itself.
> 也支持微信、QQ、企业微信和钉钉 — 见下方中文章节。
> 详细配置请直接询问 GenericAgent 本身。
---
## 🔓 Unlocking Advanced Capabilities
## 🔓 解锁高级能力
In GA, advanced capabilities are unlocked by **instructing the agent**, not by reading
docs or installing extras. Each instruction below makes GA read its pre-installed SOPs
(battle-tested playbooks in its memory), install whatever is missing, adapt to your OS,
and persist the result into its own memory.
在 GA 中，高级能力通过**向智能体下达指令**来解锁，而非阅读文档或安装额外组件。下面每条指令都会让 GA 读取其预装的 SOP（标准作业程序，Standard Operating Procedure；经实战检验、存于其记忆中的 playbook），安装缺失项、适配你的操作系统，并将结果持久化到自身记忆中。
| Capability | Just tell GA |
| :--- | :--- |
| 🌐 Web automation | *"Set up your web automation capability."* — GA guides you through the one manual step: dragging the bundled Chrome extension into `chrome://extensions`. |
| 🌐 Web automation | *"Set up your web automation capability."* — GA 会引导你完成唯一的手动步骤：将捆绑的 Chrome 扩展拖入 `chrome://extensions`。 |
| 🔤 OCR | *"Set up your OCR capability with rapidocr and save it to memory."* |
| 👁️ Vision | *"Set up your vision capability from the template in memory/."* — GA copies the template, wires it to your existing LLM keys, and self-tests. |
| 👁️ Vision | *"Set up your vision capability from the template in memory/."* — GA 会复制模板、接入你现有的 LLM 密钥并自行测试。 |
| 🖱️ Computer use | *"Probe this system and set up your computer-use capability."* |
> 💡 **About language**: the pre-installed SOPs are written in Chinese — GA reads them
> natively, so this never blocks you. If you prefer an English knowledge base, just say:
> *"Read your pre-installed SOPs and rewrite them in English (keep code, paths and error
> strings verbatim)."*
> 💡 **关于语言**：预装 SOP 以中文撰写 — GA 可原生阅读，因此不会阻碍你使用。若你希望知识库为英文，只需说：
> *"Read your pre-installed SOPs and rewrite them in English (keep code, paths and error strings verbatim)."*
>
> 🌍 **About platforms**: the SOPs were honed on Windows, but cross-platform adaptation is
> itself a GA task — on macOS/Linux, GA swaps in the platform equivalents (window
> enumeration, input control, screenshots) on its own. Same self-evolution principle.
> 🌍 **关于平台**：SOP 在 Windows 上打磨成熟，但跨平台适配本身也是 GA 的任务 — 在 macOS/Linux 上，GA 会自行替换为对应平台的等价能力（窗口枚举、输入控制、截图等）。遵循同样的自进化原则。
---
## 🧠 Architecture
## 🧠 架构
GenericAgent accomplishes complex tasks through **Layered Memory × Minimal Toolset × Autonomous Execution Loop**, continuously accumulating experience during execution.
GenericAgent 通过**分层记忆 × 极简工具集 × 自主执行循环**完成复杂任务，并在执行过程中持续积累经验。
### 1️⃣ Layered Memory System
### 1️⃣ 分层记忆系统
> *Memory crystallizes throughout task execution, letting the agent build stable, efficient working patterns over time.*
> *记忆在任务执行全程不断结晶，使智能体能够随时间建立稳定、高效的工作模式。*
| Layer | Name | Description |
| :---: | :--- | :--- |
| **L0** | Meta Rules | Core behavioral rules and system constraints |
| **L1** | Insight Index | Minimal memory index for fast routing and recall |
| **L2** | Global Facts | Stable knowledge accumulated over long-term operation |
| **L3** | Task Skills / SOPs | Reusable workflows for completing specific task types |
| **L4** | Session Archive | Archived task records distilled from finished sessions for long-horizon recall |
| **L0** | Meta Rules | 核心行为规则与系统约束 |
| **L1** | Insight Index | 用于快速路由与召回的极简记忆索引 |
| **L2** | Global Facts | 长期运行中积累的稳定知识 |
| **L3** | Task Skills / SOPs | 完成特定任务类型的可复用工作流 |
| **L4** | Session Archive | 从已完成会话中提炼并归档的任务记录，用于长程回忆 |
### 2️⃣ Autonomous Execution Loop
### 2️⃣ 自主执行循环
> *Perceive environment state → Task reasoning → Execute tools → Write experience to memory → Loop*
> *感知环境状态 → 任务推理 → 执行工具 → 将经验写入记忆 → 循环*
The entire core loop is just **~100 lines of code** ([`agent_loop.py`](agent_loop.py)).
整个核心循环仅约 **~100 行代码**（[`agent_loop.py`](agent_loop.py)）。
### 3️⃣ Minimal Toolset
### 3️⃣ 极简工具集
> *GenericAgent provides only **9 atomic tools**, forming the foundational capabilities for interacting with the outside world.*
> *GenericAgent 仅提供 **9 个原子工具**，构成与外部世界交互的基础能力。*
| Tool | Function |
| :--- | :--- |
| `code_run` | Execute arbitrary code (Python / PowerShell) |
| `file_read` | Read files |
| `file_write` | Write / create / overwrite files |
| `file_patch` | Patch / modify files |
| `web_scan` | Perceive web content |
| `web_execute_js` | Control browser behavior |
| `ask_user` | Human-in-the-loop confirmation |
| `update_working_checkpoint` | *(memory)* Short-term working notepad |
| `start_long_term_update` | *(memory)* Distill long-term memory |
| `code_run` | 执行任意代码（Python / PowerShell） |
| `file_read` | 读取文件 |
| `file_write` | 写入 / 创建 / 覆盖文件 |
| `file_patch` | 修补 / 修改文件 |
| `web_scan` | 感知网页内容 |
| `web_execute_js` | 控制浏览器行为 |
| `ask_user` | 人在回路（Human-in-the-loop）确认 |
| `update_working_checkpoint` | *（memory）* 短期工作记事本 |
| `start_long_term_update` | *（memory）* 提炼长期记忆 |
### 4️⃣ Capability Extension
### 4️⃣ 能力扩展
> *Capable of dynamically creating new tools.*
> *能够动态创建新工具。*
Via `code_run`, GenericAgent can dynamically install Python packages, write new scripts, call external APIs, or control hardware at runtime — crystallizing temporary abilities into permanent tools.
通过 `code_run`，GenericAgent 可在运行时动态安装 Python 包、编写新脚本、调用外部 API 或控制硬件 — 将临时能力结晶为永久工具。
<div align="center">
<img src="assets/images/workflow.jpg" alt="GenericAgent Workflow" width="420"/>
<br/><em>GenericAgent Workflow Diagram</em>
<br/><em>GenericAgent 工作流示意图</em>
</div>
---
## 🧬 Self-Evolution Mechanism
## 🧬 自进化机制
This is what fundamentally distinguishes GenericAgent from every other agent framework.
这正是 GenericAgent 与其他智能体框架的根本区别。
```text
[New Task]
@@ -306,16 +305,16 @@ This is what fundamentally distinguishes GenericAgent from every other agent fra
| What you say | First time | Every time after |
| :--- | :--- | :--- |
| *"Read my WeChat messages"* | Install deps → reverse DB → write read script → save Skill | **one-line invoke** |
| *"Give me a morning digest of Hacker News"* | Write scraper → build digest → schedule daily run → save Skill | **one-line invoke** |
| *"Monitor stocks and alert me"* | Install `mootdx` → build selection flow → configure cron → save Skill | **one-line start** |
| *"Send this file via Gmail"* | Configure OAuth → write send script → save Skill | **ready to use** |
| *"Read my WeChat messages"* | 安装依赖 → 逆向数据库 → 编写读取脚本 → 保存 Skill | **一行调用** |
| *"Give me a morning digest of Hacker News"* | 编写爬虫 → 构建摘要 → 安排每日运行 → 保存 Skill | **一行调用** |
| *"Monitor stocks and alert me"* | 安装 `mootdx` → 构建选股流程 → 配置 cron → 保存 Skill | **一行启动** |
| *"Send this file via Gmail"* | 配置 OAuth → 编写发送脚本 → 保存 Skill | **即可使用** |
After a few weeks, your agent instance will have a skill tree no one else in the world has — all grown from 3K lines of seed code.
几周之后，你的智能体实例将拥有一棵世界上独一无二的技能树 — 全部从 3K 行种子代码生长而来。
---
## 📊 Comparison
## 📊 对比
| Feature | **GenericAgent** | OpenClaw | Claude Code |
| :--- | :---: | :---: | :---: |
@@ -328,95 +327,95 @@ After a few weeks, your agent instance will have a skill tree no one else in the
---
## 📈 Evaluation
## 📈 评测
> 📂 Full evaluation datasets and results: [**JinyiHan99/GA-Technical-Report**](https://github.com/JinyiHan99/GA-Technical-Report/tree/main)
> 📂 完整评测数据集与结果：[**JinyiHan99/GA-Technical-Report**](https://github.com/JinyiHan99/GA-Technical-Report/tree/main)
We evaluate GenericAgent across **five dimensions**:
我们从 **五个维度** 评测 GenericAgent：
| # | Dimension | Question | Benchmarks |
| :---: | :--- | :--- | :--- |
| 1 | **Task Completion & Token Efficiency** | Can GA complete hard tasks more cheaply than leading agents? | SOP-Bench, Lifelong AgentBench, RealFin-Benchmark |
| 2 | **Tool-Use Efficiency** | Can a minimal atomic toolset solve what specialized toolsets solve, with less overhead? | Tool Efficiency Benchmark (11 simple + 5 long-horizon) |
| 3 | **Memory System Effectiveness** | Does condensed hierarchical memory beat full/redundant memory and embedding-based retrievers? | SOP-Bench (dangerous goods), LoCoMo, 20-skill stress test |
| 4 | **Self-Evolution Capability** | Can the agent distill experience into reusable SOPs and code, without intervention? | 9-round LangChain longitudinal study, 8-task cross-task web benchmark |
| 5 | **Web Browsing Capability** | Does density-driven design survive the open web? | WebCanvas, BrowseComp-ZH, Custom Tasks (22) |
| 1 | **Task Completion & Token Efficiency** | GA 能否比主流智能体更便宜地完成困难任务？ | SOP-Bench, Lifelong AgentBench, RealFin-Benchmark |
| 2 | **Tool-Use Efficiency** | 极简原子工具集能否以更低开销解决专用工具集所能解决的问题？ | Tool Efficiency Benchmark (11 simple + 5 long-horizon) |
| 3 | **Memory System Effectiveness** | 凝练的分层记忆能否优于完整/冗余记忆与基于嵌入的检索器？ | SOP-Bench (dangerous goods), LoCoMo, 20-skill stress test |
| 4 | **Self-Evolution Capability** | 智能体能否在无人工干预的情况下将经验提炼为可复用的 SOP 与代码？ | 9-round LangChain longitudinal study, 8-task cross-task web benchmark |
| 5 | **Web Browsing Capability** | 密度驱动设计能否在开放互联网上经受考验？ | WebCanvas, BrowseComp-ZH, Custom Tasks (22) |
Baselines across these dimensions include **Claude Code**, **OpenAI CodeX**, and **OpenClaw**, evaluated under *Claude Sonnet 4.6*, *Claude Opus 4.6*, *GPT-5.4*, and *MiniMax M2.7* backbones.
各维度的基线包括 **Claude Code**、**OpenAI CodeX** 和 **OpenClaw**，在 *Claude Sonnet 4.6*、*Claude Opus 4.6*、*GPT-5.4* 与 *MiniMax M2.7* 骨干模型下评测。
<table>
<tr>
<td align="center" width="50%">
<img src="assets/images/result_radar.png" width="100%" alt="Tool-use efficiency radar"/><br/>
<sub><b>Tool-use efficiency radar.</b> GA dominates token, request, and tool-call axes while preserving quality across four task dimensions.</sub>
<sub><b>工具使用效率雷达图。</b> GA 在 token、请求与工具调用轴上占主导，同时在四个任务维度上保持质量。</sub>
</td>
<td align="center" width="50%">
<img src="assets/images/result_convergence.png" width="100%" alt="Cross-task self-evolution convergence"/><br/>
<sub><b>Cross-task self-evolution.</b> Second- and third-run GA executions converge to a stable low-cost regime across eight web tasks, while OpenClaw shows no such convergence.</sub>
<sub><b>跨任务自进化。</b> GA 第二、第三次运行在八个网页任务上收敛至稳定的低成本区间，而 OpenClaw 未出现此类收敛。</sub>
</td>
</tr>
</table>
### Browser Realness of GA Web Tools (TMWebdriver)
### GA Web Tools 的浏览器真实性（TMWebdriver）
GA web tools are powered by **TMWebdriver** — a local WebSocket server plus a Chrome extension — running through a **real, persistent Chrome/Chromium session** rather than a disposable headless sandbox, preserving cookies, login state, extensions, GPU/WebGL behavior, and normal browser-session fingerprints.
GA Web Tools 由 **TMWebdriver** 驱动 —— 本地 WebSocket 服务器加 Chrome 扩展 —— 在**真实、持久的 Chrome/Chromium 会话**中运行，而非一次性无头沙箱，从而保留 cookies、登录状态、扩展、GPU/WebGL 行为以及正常的浏览器会话指纹。
| Detection Service / Signal | Vanilla Headless Automation | GA Web Tools | Notes |
| 检测服务 / 信号 | 普通无头自动化 | GA Web Tools | 备注 |
| :--- | :---: | :---: | :--- |
| SannySoft headless test | Often detected | ✅ 56/56 passed | `bot.sannysoft.com` |
| bot.incolumitas.com | Commonly fails webdriver / CDP checks | ✅ 36/36 passed | `WEBDRIVER`, `SELENIUM_DRIVER`, `webDriverAdvanced` all OK |
| BrowserScan bot detection | Often abnormal | ✅ Normal | `browserscan.net` |
| Device & Browser Info bot test | Multiple bot flags | ✅ Human / `isBot=false` | `deviceandbrowserinfo.com` |
| FingerprintJS bot detection demo | Often detected | ✅ Passed | Demo flow completed without bot verdict |
| reCAPTCHA v3 demo | Low bot-like score | ✅ 0.9 human-like score | Score-based risk signal; 0.9 is above typical production thresholds |
| SannySoft 无头测试 | 常被检测到 | ✅ 56/56 通过 | `bot.sannysoft.com` |
| bot.incolumitas.com | 常在 webdriver / CDP 检查中失败 | ✅ 36/36 通过 | `WEBDRIVER`、`SELENIUM_DRIVER`、`webDriverAdvanced` 均正常 |
| BrowserScan 机器人检测 | 常显示异常 | ✅ 正常 | `browserscan.net` |
| Device & Browser Info 机器人测试 | 多项机器人标记 | ✅ 人类 / `isBot=false` | `deviceandbrowserinfo.com` |
| FingerprintJS 机器人检测演示 | 常被检测到 | ✅ 通过 | 演示流程完成，无机器人判定 |
| reCAPTCHA v3 演示 | 低机器人相似度分数 | ✅ 0.9 人类相似度分数 | 基于分数的风险信号；0.9 高于典型生产环境阈值 |
For reCAPTCHA v3, `0.9` is not a "checkbox solved" result; it is the high-confidence human-like score returned by the risk model, typically sufficient to avoid extra challenges in production flows.
对于 reCAPTCHA v3，`0.9` 并非「复选框已解决」的结果；它是风险模型返回的高置信度人类相似度分数，通常足以在生产流程中避免额外验证挑战。
---
## 📅 Roadmap & News
## 📅 路线图与最新动态
- **2026-05-23** — 🆕 **TUI v3 released** (`frontends/tui_v3.py`). Block-based scrollback with proper resize reflow, per-terminal color profile for cross-terminal parity, and feature parity with v2.
- **2026-05-18** — 🆕 **Morphling mode**. Project-level skill absorption — extract goal + tests from any external repo, then decide per component: call, rewrite, or discard. See `memory/morphling_sop.md`.
- **2026-05-17** — 🆕 **Goal Hive mode**. Multi-worker cooperative Goal mode — BBS-coordinated master/workers running long-horizon objectives in parallel. See `memory/goal_hive_sop.md`.
- **2026-05-15** — 🖥️ **Desktop GUI released**. One-line installs ship a ready-to-run desktop app (`frontends/GenericAgent.exe`). Developers launch via `python launch.pyw`.
- **2026-05-14** — 🆕 **Conductor sub-agent orchestration**. Spawn, supervise, and auto-clean parallel sub-agents; first-class delegation primitives complementing `/btw` side-questions.
- **2026-05-12** — 🆕 **TUI v2 released** (`frontends/tuiapp_v2.py`). Refined Textual frontend with image-paste folding, file paste, block-delete, Ctrl+C copy, history navigation, and `/llm` / `/export` / `/continue` pickers.
- **2026-05-08** — 🆕 **Goal mode** (`reflect/goal_mode.py`). Time-budget-driven self-driven loop — "keep optimizing X for N hours" with no premature delivery.
- **2026-04-21** — 📄 [**Technical Report on arXiv**](https://arxiv.org/abs/2604.17091) — *GenericAgent: A Token-Efficient Self-Evolving LLM Agent via Contextual Information Density Maximization*.
- **2026-04-11** — Introduced **L4 session archive memory** and scheduler cron integration.
- **2026-03-23** — Personal WeChat supported as a bot frontend.
- **2026-03-10** — [Released million-scale Skill Library](https://mp.weixin.qq.com/s/q2gQ7YvWoiAcwxzaiwpuiQ?scene=1&click_id=7) *(Chinese)*.
- **2026-03-08** — [Released "Dintal Claw" — a GenericAgent-powered government-affairs bot](https://mp.weixin.qq.com/s/eiEhwo-j6S-WpLxgBnNxBg) *(Chinese)*.
- **2026-03-01** — [Featured by Jiqizhixin (机器之心)](https://mp.weixin.qq.com/s/uVWpTTF5I1yzAENV_qm7yg) *(Chinese)*.
- **2026-01-16** — GenericAgent **V1.0** public release.
- **2026-05-23** — 🆕 **TUI v3 发布**（`frontends/tui_v3.py`）。基于块的滚动回溯、正确的调整大小重排、用于跨终端一致性的每终端配色方案，以及与 v2 的功能对等。
- **2026-05-18** — 🆕 **Morphling 模式**。项目级技能吸收 —— 从任意外部仓库提取目标与测试，再按组件决定：调用、重写或丢弃。详见 `memory/morphling_sop.md`。
- **2026-05-17** — 🆕 **Goal Hive 模式**。多 Worker 协作式 Goal 模式 —— 由 BBS 协调的 master/workers 并行运行长期目标。详见 `memory/goal_hive_sop.md`。
- **2026-05-15** — 🖥️ **桌面 GUI 发布**。一行安装即可交付开箱即用的桌面应用（`frontends/GenericAgent.exe`）。开发者通过 `python launch.pyw` 启动。
- **2026-05-14** — 🆕 **Conductor 子 Agent 编排**。生成、监督并自动清理并行子 Agent；一等委托原语，补充 `/btw` 侧问题。
- **2026-05-12** — 🆕 **TUI v2 发布**（`frontends/tuiapp_v2.py`）。精炼的 Textual 前端，支持图片粘贴折叠、文件粘贴、块删除、Ctrl+C 复制、历史导航，以及 `/llm` / `/export` / `/continue` 选择器。
- **2026-05-08** — 🆕 **Goal 模式**（`reflect/goal_mode.py`）。时间预算驱动的自主循环 ——「在 N 小时内持续优化 X」，不会提前交付。
- **2026-04-21** — 📄 [**arXiv 技术报告**](https://arxiv.org/abs/2604.17091) — *GenericAgent: A Token-Efficient Self-Evolving LLM Agent via Contextual Information Density Maximization*。
- **2026-04-11** — 引入 **L4 会话归档记忆**与调度器 cron 集成。
- **2026-03-23** — 支持个人微信作为机器人前端。
- **2026-03-10** — [发布百万级 Skill 库](https://mp.weixin.qq.com/s/q2gQ7YvWoiAcwxzaiwpuiQ?scene=1&click_id=7) *(Chinese)*.
- **2026-03-08** — [发布「Dintal Claw」—— 由 GenericAgent 驱动的政务机器人](https://mp.weixin.qq.com/s/eiEhwo-j6S-WpLxgBnNxBg) *(Chinese)*.
- **2026-03-01** — [获机器之心（Jiqizhixin）报道](https://mp.weixin.qq.com/s/uVWpTTF5I1yzAENV_qm7yg) *(Chinese)*.
- **2026-01-16** — GenericAgent **V1.0** 公开发布。
---
## ⭐ Community & Support
## ⭐ 社区与支持
If this project helped you, please consider leaving a **Star!** 🙏
如果本项目对你有帮助，欢迎点个 **Star！** 🙏
### 🚩 Friendly Links
### 🚩 友情链接
Thanks to the **LinuxDo** community for the support!
感谢 **LinuxDo** 社区的支持！
[![LinuxDo](https://img.shields.io/badge/Community-LinuxDo-blue?style=for-the-badge)](https://linux.do/)
**Community GUIs** *(independent open-source projects)*:
**社区 GUI** *(独立开源项目)*：
- [chilishark27/ga-manager](https://github.com/chilishark27/ga-manager)
- [wangjc683/galley](https://github.com/wangjc683/galley) — Out-of-the-box local agent workbench with a bundled GA runtime (CPython 3.11 + deps), native GUI/CLI, multi-session + Project orchestration, local-first.
- [wangjc683/galley](https://github.com/wangjc683/galley) — 开箱即用的本地 Agent 工作台，捆绑 GA 运行时（CPython 3.11 + 依赖），原生 GUI/CLI，多会话 + Project 编排，本地优先。
- [FroStorM/A3Agent](https://github.com/FroStorM/A3Agent/tree/workbench)
- [Fwind43/GenericAgent-Admin](https://github.com/Fwind43/GenericAgent-Admin) — Go + React desktop admin panel: service lifecycle management, native chat, Goal mode, BBS team board, file editor, model config wizard, TMWebDriver monitor, self-update, and Windows tray/desktop-pet integration.
- [Fwind43/GenericAgent-Admin](https://github.com/Fwind43/GenericAgent-Admin) — Go + React 桌面管理面板：服务生命周期管理、原生聊天、Goal 模式、BBS 团队看板、文件编辑器、模型配置向导、TMWebDriver 监控、自更新，以及 Windows 托盘/桌面宠物集成。
---
## 📄 License
## 📄 许可
Distributed under the **MIT License**. See [`LICENSE`](LICENSE) for full text.
基于 **MIT License** 分发。完整文本见 [`LICENSE`](LICENSE)。
> *Disclaimer: The official GenericAgent channels are this GitHub repository and https://gaagent.ai. DintalClaw is currently the only officially authorized commercial partner; any other third-party website, organization, or individual using the GenericAgent name is not official unless explicitly listed here.*
> *免责声明：GenericAgent 官方渠道为本 GitHub 仓库与 https://gaagent.ai.；DintalClaw 是目前唯一官方授权商业合作伙伴；任何其他使用 GenericAgent 名称的第三方网站、组织或个人，除非在此明确列出，否则均非官方。*
---
@@ -802,7 +801,7 @@ GA Web 工具运行在**真实、持久化的 Chrome/Chromium 会话**中，而
---
## 📈 Star History
## 📈 Star 历史
<div align="center">
@@ -810,7 +809,7 @@ GA Web 工具运行在**真实、持久化的 Chrome/Chromium 会话**中，而
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=lsdefine/GenericAgent&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=lsdefine/GenericAgent&type=Date" />
<img alt="Star History Chart" src="https://api.star-history.com/svg?repos=lsdefine/GenericAgent&type=Date" />
<img alt="Star 历史图表" src="https://api.star-history.com/svg?repos=lsdefine/GenericAgent&type=Date" />
</picture>
</a>