chore(docs): reorganize docs/ tree + rename clawspring → cheetahclaws…

docs/README.md

@@ -0,0 +1,29 @@
+# CheetahClaws Documentation Index
+
+This directory holds all project documentation. For the project overview,
+start with the [top-level README](../README.md).
+
+## Layout
+
+| Path | What |
+|------|------|
+| [`news.md`](news.md) | Full release notes (the top-level README has one-line summaries linking here) |
+| [`agent-os.md`](agent-os.md) | Agent-OS layer (`cc_kernel/`) reference |
+| [`architecture.md`](architecture.md) | System architecture deep dive |
+| [`contributor_guide.md`](contributor_guide.md) | How to set up dev environment and submit PRs |
+| [`guides/`](guides/) | User-facing how-to guides (Web UI, bridges, trading, research lab, plugin authoring, voice/video, recipes, reference, Docker, advanced) |
+| [`RFC/`](RFC/) | Design proposals (numbered, ~30 documents covering daemon, kernel, agent OS) |
+| [`roadmap/`](roadmap/) | Roadmap snapshots |
+| [`superpowers/`](superpowers/) | Plans and specs for in-flight initiatives |
+| [`PR/`](PR/) | Per-PR design notes |
+| [`media/`](media/) | Images, logos, demo GIFs (`demos/`, `logos/`, `screenshots/`) |
+| [`i18n/`](i18n/) | README translations (CN · DE · ES · FR · JP · KO · PT) |
+| [`archive/`](archive/) | Superseded docs (old release notes, old comparison tables) |
+
+## Where to start
+
+- **Just want to use it?** Start with the [top-level README](../README.md), then [`guides/reference.md`](guides/reference.md) for the full slash-command and config reference.
+- **Setting up the Web UI?** [`guides/web-ui.md`](guides/web-ui.md).
+- **Building a plugin / tool?** [`guides/plugin-authoring.md`](guides/plugin-authoring.md).
+- **Hacking on the core?** [`architecture.md`](architecture.md) + [`contributor_guide.md`](contributor_guide.md).
+- **Curious about the agent-OS direction?** [`agent-os.md`](agent-os.md) + the `RFC/` series (start with `0002-daemon-foundation-roadmap.md`).

docs/guides/advanced.md

@@ -69,7 +69,7 @@ Generating diverse perspectives...
 `/ssj` opens a persistent interactive power menu — a single entry point for the most common development workflows, so you never have to remember command names.
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/ssj_demo.gif" width="850"/>
+<img src="../media/demos/ssj_demo.gif" width="850"/>
 </div>
 
 ### Menu options
@@ -146,7 +146,7 @@ Any `/command` typed at the `⚡ SSJ »` prompt is passed through to the REPL:
 ### Worker command
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/worker_demo.gif" width="850"/>
+<img src="../media/demos/worker_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">/worker: brainstorm → 5-task queue → auto-implement each with progress bar</center>
@@ -218,7 +218,7 @@ SSJ option 4 runs a structured multi-round expert debate on any file:
 ## Tmux Integration
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/tmux_demo.gif" width="850"/>
+<img src="../media/demos/tmux_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Tmux Integration: AI splits panes, sends commands, captures output across sessions</center>
@@ -288,7 +288,7 @@ The AI calls `TmuxNewSession(detached=true)` then sends the vLLM launch command
 ## Shell Escape
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/shell_escape_demo.gif" width="850"/>
+<img src="../media/demos/shell_escape_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Shell Escape: ! prefix runs commands directly — git, ls, python, pipes — no AI involvement</center>
@@ -370,7 +370,7 @@ Duration suffix: `s` = seconds, `m` = minutes, `h` = hours. Plain integer = seco
 ## Checkpoint System
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/checkpoint_demo.gif" width="850"/>
+<img src="../media/demos/checkpoint_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Checkpoint / Rewind: AI breaks tests → /checkpoint list → rewind → files restored</center>
@@ -416,7 +416,7 @@ CheetahClaws automatically snapshots your conversation and any edited files afte
 ## Plan Mode
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/plan_demo.gif" width="850"/>
+<img src="../media/demos/plan_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Plan Mode: Read-only analysis → write plan → /plan done → implement</center>
@@ -681,7 +681,7 @@ Resume a specific file:
 ## Cloud Sync (GitHub Gist)
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/cloudsave_demo.gif" width="850"/>
+<img src="../media/demos/cloudsave_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Cloud Sync: /cloudsave on desktop → encrypted upload → /cloudload on laptop → full session restored</center>

docs/guides/bridges.md

@@ -5,7 +5,7 @@
 `/telegram` turns cheetahclaws into a Telegram bot — receive messages from your phone, run the model with full tool access, and reply automatically.
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/telegram_demo.gif" width="850"/>
+<img src="../media/demos/telegram_demo.gif" width="850"/>
 </div>
 
 ### Setup (one-time)
@@ -188,7 +188,7 @@ The bridge also auto-starts in **web-server mode** (`cheetahclaws --web`) — ha
 ## WeChat Bridge
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/wechat_demo.gif" width="850"/>
+<img src="../media/demos/wechat_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">WeChat Bridge: Control cheetahclaws from WeChat (微信)</center>
@@ -398,7 +398,7 @@ stdout, you copy/paste into your IM client of choice.
 ## Slack Bridge
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/slack_demo.gif" width="850"/>
+<img src="../media/demos/slack_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Slack Bridge: Control cheetahclaws from Slack</center>

docs/guides/extensions.md

@@ -3,7 +3,7 @@
 ## Memory
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/memory_demo.gif" width="850"/>
+<img src="../media/demos/memory_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Memory: save preferences in session 1 → auto-recalled in session 2, no re-explanation needed</center>
@@ -191,7 +191,7 @@ AI: [deploys version 2.1.0 to staging]
 ## Sub-Agents
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/subagent_demo.gif" width="850"/>
+<img src="../media/demos/subagent_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Sub-Agents: spawn coder + security agents in parallel, merge results automatically</center>

docs/guides/reference.md

@@ -160,7 +160,7 @@ Type `/` and press **Tab** to see all commands with descriptions. Continue typin
 **Switching models inside a session:**
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/multimodel_demo.gif" width="850"/>
+<img src="../media/demos/multimodel_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Multi-Model Switching: Claude → GPT-4o → Ollama → back, full history preserved</center>

docs/guides/trading.md

@@ -1,7 +1,7 @@
 # Trading Agent
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/trading_demo.gif" width="850"/>
+<img src="../media/demos/trading_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Trading Agent: SSJ → multi-agent analysis (Bull/Bear debate + Risk panel + PM decision) → backtest → indicators</center>

docs/guides/voice-and-video.md

@@ -3,7 +3,7 @@
 ## Voice Input
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/voice_demo.gif" width="850"/>
+<img src="../media/demos/voice_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Voice Input: speak your prompt → offline Whisper transcription → AI responds</center>
@@ -105,7 +105,7 @@ The selected device is shown in `/voice status` and used for all subsequent reco
 ## Video Content Factory
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/video_demo.gif" width="850"/>
+<img src="../media/demos/video_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Video Factory: topic → AI story → TTS → images → subtitles → final .mp4</center>
@@ -205,7 +205,7 @@ Subtitles are rendered with **Pillow + NotoSansSC font** — not libass. This me
 ### Image backends (vision input)
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/image_demo.gif" width="850"/>
+<img src="../media/demos/image_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">/image: paste UI screenshot → AI flags issues; paste code screenshot → AI spots bugs</center>
@@ -306,7 +306,7 @@ video_tmp/batch_20260407_153000/story/
 ## TTS Content Factory
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/voice_demo.gif" width="850"/>
+<img src="../media/demos/voice_demo.gif" width="850"/>
 </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">TTS Factory: choose voice style → AI writes script → synthesize → .mp3 output</center>

docs/guides/web-ui.md

@@ -3,7 +3,7 @@
 CheetahClaws ships with a production-ready browser UI built on a pure Python stdlib HTTP server plus nine small vanilla-JS modules — no Node.js, no bundler, no build step. This guide covers installation, accounts, the Chat UI, the PTY terminal, the full HTTP API, observability, and how the pieces fit together.
 
 <div align="center">
- <img src="https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/web_demo.gif" width="850" alt="Web UI demo — sidebar, tool cards, approval prompts, markdown streaming"/>
+ <img src="../media/demos/web_demo.gif" width="850" alt="Web UI demo — sidebar, tool cards, approval prompts, markdown streaming"/>
 </div>
 
 ---

docs/README.CN.MD → docs/i18n/README.CN.MD

@@ -1,26 +1,26 @@
-English | [中文](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/README.CN.MD) | [Français](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/README.FR.MD) | [한국어](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/README.KO.MD) | [日本語](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/README.JP.MD) | [Deutsch](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/README.DE.MD) | [Português](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/README.ES.MD)
+English | [中文](README.CN.MD) | [Français](README.FR.MD) | [한국어](README.KO.MD) | [日本語](README.JP.MD) | [Deutsch](README.DE.MD) | [Português](README.ES.MD)
 
 <br>
 
 <div align="center">
-  <a href="[https://github.com/SafeRL-Lab/Robust-Gymnasium](https://github.com/SafeRL-Lab/clawspring)">
-    <img src="logo-5.png" alt="Logo" width="280"> 
+  <a href="[https://github.com/SafeRL-Lab/Robust-Gymnasium](https://github.com/SafeRL-Lab/cheetahclaws)">
+    <img src="../media/logos/logo-5.png" alt="Logo" width="280"> 
   </a>
 
 
 <h2 align="center" style="font-size: 30px;"><strong><em>CheetahClaws (代码猎豹)</em></strong>：一个快速、易用、原生基于 Python 的任意模型个人 AI 助手，灵感来自 OpenClaw 与 Claude Code，旨在为你 24/7 自主工作</h2>
 <p align="center">
     <a href="https://github.com/chauncygu/collection-claude-code-source-code">Claude Code 最新源码集合</a>
     ·
-    <a href="https://github.com/SafeRL-Lab/clawspring/issues">Issue</a>
+    <a href="https://github.com/SafeRL-Lab/cheetahclaws/issues">Issue</a>
   ·
-    <a href="https://deepwiki.com/SafeRL-Lab/clawspring">简要介绍</a>
+    <a href="https://deepwiki.com/SafeRL-Lab/cheetahclaws">简要介绍</a>
 
   </p>
 </div>
 
  <div align=center>
- <img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/demo.gif" width="850"/> 
+ <img src="../media/demos/demo.gif" width="850"/> 
  </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">任务执行</center>
@@ -30,7 +30,7 @@ English | [中文](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/READM
 ---
 
   <div align=center>
- <img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/brainstorm_demo.gif" width="850"/> 
+ <img src="../media/demos/brainstorm_demo.gif" width="850"/> 
  </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">头脑风暴模式：多智能体头脑风暴</center>
@@ -41,7 +41,7 @@ English | [中文](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/READM
 ---
 
   <div align=center>
- <img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/proactive_demo.gif" width="850"/> 
+ <img src="../media/demos/proactive_demo.gif" width="850"/> 
  </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">主动模式：自主 Agent</center>
@@ -50,7 +50,7 @@ English | [中文](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/READM
 ---
 
   <div align=center>
- <img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/ssj_demo.gif" width="850"/> 
+ <img src="../media/demos/ssj_demo.gif" width="850"/> 
  </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">SSJ 开发者模式：强力菜单工作流</center>
@@ -59,7 +59,7 @@ English | [中文](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/READM
 ---
 
   <div align=center>
- <img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/telegram_demo.gif" width="850"/> 
+ <img src="../media/demos/telegram_demo.gif" width="850"/> 
  </div>
 <div align=center>
 <center style="color:#000000;text-decoration:underline">Telegram 桥接：用手机控制 cheetahclaws</center>
@@ -116,7 +116,7 @@ English | [中文](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/READM
   - 所有文档、GitHub URL 和内部引用均已同步更新。
   - README 中新增 **CheetahClaws vs OpenClaw** 对比章节。
 
-更多更新见 [这里](https://github.com/SafeRL-Lab/cheetahclaws/blob/main/docs/news.md)
+更多更新见 [这里](../news.md)
 
 
 ---
@@ -170,7 +170,7 @@ CheetahClaws：**一个轻量级**、**易于使用**、支持**任意模型**
 
 Claude Code 是一个强大、面向生产环境的 AI 编程助手 —— 但它的源码是一个编译后的 12 MB TypeScript/Node.js bundle（约 1,300 个文件，约 283K 行代码）。它与 Anthropic API 紧密耦合，难以修改，也无法运行在本地模型或其他替代模型之上。
 
-**CheetahClaws** 用约 10K 行可读的 Python 代码重实现了相同的核心循环，保留你需要的部分，去掉你不需要的部分。更详细分析见此（CheetahClaws v3.03），[英文版](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/comparison_claude_code_vs_nano_v3.03_en.md) 与 [中文版](https://github.com/SafeRL-Lab/clawspring/blob/main/docs/comparison_claude_code_vs_nano_v3.03_cn.md)
+**CheetahClaws** 用约 10K 行可读的 Python 代码重实现了相同的核心循环，保留你需要的部分，去掉你不需要的部分。更详细分析见此（CheetahClaws v3.03），[英文版](../comparison_claude_code_vs_nano_v3.03_en.md) 与 [中文版](../comparison_claude_code_vs_nano_v3.03_cn.md)
 
 ### 快速概览
 
@@ -419,7 +419,7 @@ Claude Code 是一个强大、面向生产环境的 AI 编程助手 —— 但
 curl -LsSf https://astral.sh/uv/install.sh | sh
 
 # 克隆并安装
-git clone https://github.com/SafeRL-Lab/clawspring
+git clone https://github.com/SafeRL-Lab/cheetahclaws
 cd cheetahclaws
 uv tool install .
 ```
@@ -447,7 +447,7 @@ uv tool uninstall cheetahclaws
 ### 备选方案：直接从仓库运行
 
 ```bash
-git clone https://github.com/SafeRL-Lab/clawspring
+git clone https://github.com/SafeRL-Lab/cheetahclaws
 cd cheetahclaws
 
 pip install -r requirements.txt
@@ -1685,7 +1685,7 @@ Generating diverse perspectives...
 `/ssj` 会打开一个持久化交互强力菜单 —— 统一入口覆盖最常见的开发工作流，这样你不再需要记命令名。
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/ssj_demo.gif" width="850"/>
+<img src="../media/demos/ssj_demo.gif" width="850"/>
 </div>
 
 ### 菜单项
@@ -1820,7 +1820,7 @@ SSJ 菜单第 4 项会对任意文件运行结构化多轮专家辩论：
 `/telegram` 会将 cheetahclaws 变成一个 Telegram bot —— 从手机接收消息、以完整工具能力运行模型并自动回复。
 
 <div align=center>
-<img src="https://github.com/SafeRL-Lab/clawspring/blob/main/docs/telegram_demo.gif" width="850"/>
+<img src="../media/demos/telegram_demo.gif" width="850"/>
 </div>
 
 ### 设置（一次性）