docs: update README with bilingual version and evolutionary intelligence description
This commit is contained in:
Liang Jiaqing
117
README.md
117
README.md
@@ -1,42 +1,105 @@
|
|||||||
# pc-agent-loop
|
# PC-Agent-Loop: High-Performance Autonomous PC Controller
|
||||||
|
|
||||||
|
[English](#english) | [中文说明](#chinese)
|
||||||
|
|
||||||
|
<a name="english"></a>
|
||||||
|
|
||||||
|
PC-Agent-Loop is a minimalist yet powerful autonomous agent framework designed to bridge Large Language Models with direct OS-level execution. Unlike traditional chatbots, it possesses "physical" agency—the ability to perceive its environment, reason about complex goals, and execute multi-step operations across the file system, browsers, and local applications.
|
||||||
|
|
||||||
|
## 🚀 Evolutionary Intelligence & Extensibility
|
||||||
|
|
||||||
|
This agent is not limited to a fixed set of features. Its true power lies in its ability to **autonomously discover environment-specific capabilities** and **manufacture its own tools**:
|
||||||
|
|
||||||
|
- **Self-Discovery via Long-Term Memory**:
|
||||||
|
- The agent maintains a "Global Memory" (L2 Facts) to store system paths, credentials, and environmental status.
|
||||||
|
- It can autonomously retrieve context-aware SOPs (Standard Operating Procedures) to handle specialized tasks like WeChat database decryption or Gmail API operations.
|
||||||
|
- **Dynamic Tool Manufacturing**:
|
||||||
|
- Through `code_run`, the agent can write and execute arbitrary Python scripts to interface with new hardware or software.
|
||||||
|
- Examples of self-integrated capabilities include:
|
||||||
|
- **Deep Web Interaction**: JS injection via Tampermonkey for UI automation.
|
||||||
|
- **Digital Forensics**: Querying SQLCipher-encrypted databases (e.g., WeChat v4.0+).
|
||||||
|
- **Vision-Driven Logic**: Understanding UI states through local vision APIs (`ask_vision`).
|
||||||
|
- **System Indexing**: Utilizing **Everything CLI (es.exe)** for instant file discovery across the entire OS.
|
||||||
|
- **Android Automation**: ADB-based control for mobile device interaction.
|
||||||
|
|
||||||
|
## 📂 Project Architecture
|
||||||
|
|
||||||
|
- `agent_loop.py`: The core "Sense-Think-Act" engine (under 100 lines) driving the autonomous cycle.
|
||||||
|
- `ga.py`: The fundamental atomic toolset (File, Web, Code, User interaction).
|
||||||
|
- `agentapp.py` & `launch.pyw`: A Streamlit-based graphical interface and persistent launcher.
|
||||||
|
- `sidercall.py`: Robust LLM session management supporting multiple backends and model switching.
|
||||||
|
|
||||||
|
## 🛠️ Usage Examples
|
||||||
|
|
||||||
|
### 1. Autonomous Environment Adaptation
|
||||||
|
"Scan my local memory for recent SOPs regarding mail processing, then find and download my latest reimbursement receipts from Gmail."
|
||||||
|
|
||||||
|
### 2. Complex Multi-Step Automation
|
||||||
|
"Locate the WeChat database, decrypt it to find messages about 'Project X', and summarize the findings into a PDF report."
|
||||||
|
|
||||||
|
### 3. Real-Time System Intervention
|
||||||
|
"Monitor my cloud dashboard via the browser; if the status turns red, execute a local PowerShell script to restart the service and notify me."
|
||||||
|
|
||||||
|
## 🧩 Atomic Toolset (The Primitives)
|
||||||
|
|
||||||
|
The agent achieves high-level goals by orchestrating these 7 primitive actions:
|
||||||
|
1. `code_run`: The ultimate "Swiss Army Knife" for executing Python/PowerShell.
|
||||||
|
2. `web_scan`: Semantic perception of live web pages and tabs.
|
||||||
|
3. `web_execute_js`: Direct physical interaction with web DOM elements.
|
||||||
|
4. `file_read` & `file_write`: Direct disk access and file management.
|
||||||
|
5. `file_patch`: Safe, block-level code modification to evolve its own scripts.
|
||||||
|
6. `ask_user`: Bridging the gap for human decision-making or sensitive credentials.
|
||||||
|
7. `conclude_and_reflect`: The mechanism for distilling experiences into long-term memory.
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
<a name="chinese"></a>
|
||||||
|
|
||||||
|
# PC-Agent-Loop: 高性能 PC 级自主 AI Agent
|
||||||
|
|
||||||
pc-agent-loop 是一个极致简约的 PC 级自主 AI Agent 框架。它通过不到 100 行的核心引擎代码，构筑了对浏览器、终端和文件系统的物理级自动化能力。
|
pc-agent-loop 是一个极致简约的 PC 级自主 AI Agent 框架。它通过不到 100 行的核心引擎代码，构筑了对浏览器、终端和文件系统的物理级自动化能力。
|
||||||
|
|
||||||
## 🚀 核心特性
|
## 🚀 进化智能与扩展性
|
||||||
|
|
||||||
- **极简设计**: 仅由 7 个原子工具和一个高效的 Agentic Loop 构成。
|
本 Agent 不局限于预设功能。其核心优势在于能够**自主发现环境特定能力**并**制造属于自己的工具**：
|
||||||
- **自主代码执行**: 能够根据任务需求自主编写并运行 Python 或 PowerShell 脚本，直接操控系统资源。
|
|
||||||
- **深度 Web 自动化**: 提供语义化网页扫描与 JS 注入执行，实现精准的浏览器控制。
|
- **基于长期记忆的自我发现**:
|
||||||
- **精准文件编辑**: 支持基于源码块匹配的 `file_patch` 功能。
|
- Agent 维护“全局记忆”（L2 Facts）以存储系统路径、凭据和环境状态。
|
||||||
- **人机协作**: 在关键决策点主动请求人类干预。
|
- 能够自主检索上下文相关的 SOP（标准作业程序），以处理微信数据库解密、Gmail API 操作等专业任务。
|
||||||
|
- **动态工具制造**:
|
||||||
|
- 通过 `code_run`，Agent 可以编写并执行 Python/PowerShell 脚本来对接新硬件或软件。
|
||||||
|
- **自集成能力示例**:
|
||||||
|
- **深度 Web 自动化**: 通过 Tampermonkey 进行 JS 注入实现 UI 自动化。
|
||||||
|
- **数字取证**: 查询 SQLCipher 加密的数据库（如微信 v4.0+）。
|
||||||
|
- **视觉驱动逻辑**: 通过本地视觉 API (`ask_vision`) 理解 UI 状态。
|
||||||
|
- **系统全盘索引**: 利用 **Everything CLI (es.exe)** 实现毫秒级文件检索。
|
||||||
|
- **安卓自动化**: 基于 ADB 控制移动设备交互。
|
||||||
|
|
||||||
## 📂 项目结构
|
## 📂 项目结构
|
||||||
|
|
||||||
- `agent_loop.py`: 核心引擎，负责“感知-思考-行动”的自主循环逻辑。
|
- `agent_loop.py`: 核心引擎，负责“感知-思考-行动”的自主循环逻辑。
|
||||||
- `ga.py`: 工具箱，定义了 7 大原子工具的具体实现。
|
- `ga.py`: 工具箱，定义了原子工具的具体实现。
|
||||||
- `agentapp.py`: 基于 Streamlit 构建的交互式 Web 界面。
|
- `agentapp.py` & `launch.pyw`: 基于 Streamlit 的交互界面与持久化启动器。
|
||||||
- `sidercall.py`: LLM 通信层，支持流式输出与 API 调用。
|
- `sidercall.py`: LLM 通信层，支持多后端切换。
|
||||||
|
|
||||||
## 🛠️ 如何启动
|
## 🛠️ 典型使用场景
|
||||||
|
|
||||||
为了使 Agent 正常工作，你需要进行以下手动配置：
|
1. **环境自适应**: “扫描我的本地记忆寻找邮件处理 SOP，然后从 Gmail 下载最新的报销收据。”
|
||||||
|
2. **跨模块协作**: “定位微信数据库并解密，查找关于‘项目 X’的消息，并汇总成 PDF 报告。”
|
||||||
|
3. **系统干预**: “监控云端控制台，若状态异常则执行本地脚本重启服务并邮件通知我。”
|
||||||
|
|
||||||
1. **API Key 设置**: 在 `sidercall.py` 中设置你的 LLM API 访问 Key。
|
## 🧩 7 大核心原子工具
|
||||||
2. **Session 修改**: 在 `agentapp.py` 的 `init` 方法中，根据需要修改使用的 `LLMSession` 实例。
|
|
||||||
|
|
||||||
配置完成后，在项目根目录下执行：
|
1. `code_run`: 终极工具，执行 Python/PowerShell 脚本。
|
||||||
```bash
|
2. `web_scan`: 网页与标签页的语义化感知。
|
||||||
python launch.pyw
|
3. `web_execute_js`: 物理级网页操控（点击、滚动、数据提取）。
|
||||||
```
|
4. `file_read` & `file_write`: 磁盘文件直接访问。
|
||||||
|
5. `file_patch`: 安全的源码级局部修改。
|
||||||
|
6. `ask_user`: 关键决策或凭据输入时的人机协作。
|
||||||
|
7. `conclude_and_reflect`: 将执行经验提炼进长期记忆的机制。
|
||||||
|
|
||||||
## 🧩 7 大核心工具
|
## ⚠️ 警告
|
||||||
|
本 Agent 具备执行本地代码和控制操作系统的**物理权限**。请务必在受信任的环境中运行。
|
||||||
Agent 仅依靠以下原子工具的组合来完成任务：
|
|
||||||
`code_run`, `web_scan`, `web_execute_js`, `file_read`, `file_write`, `file_patch`, `ask_user`。
|
|
||||||
|
|
||||||
---
|
---
|
||||||
|
*Note: This README was autonomously generated and refined by the Agent.*
|
||||||
### 📝 自动生成说明
|
|
||||||
**特别说明**：本 `README.md` 文件、项目中的核心 Prompt 以及工具描述（Tools SCHEMA）完全由 Agent 自主生成并迭代优化。
|
|
||||||
|
|
||||||
**⚠️ 警告**: 本 Agent 具备执行本地代码和控制操作系统的物理权限。请务必在受信任的环境中运行。
|
|
||||||
Reference in New Issue
Block a user