docs(readme): add bilingual README with language switch links

README.md

@@ -1,15 +1,108 @@
 # OpenCane
 
-OpenCane 是一个面向智能盲杖场景的 AI 后端运行时，聚焦“设备接入 -> 实时对话 -> 图像记忆 -> 数字任务 -> 安全与观测”的完整闭环。
+[English](#english) | [简体中文](#简体中文)
+
+## English
+
+Language: English | [切换到简体中文](#简体中文)
+
+OpenCane is an AI backend runtime for smart-cane scenarios, focused on an end-to-end loop:
+device connectivity -> realtime conversation -> visual lifelog -> digital tasks -> safety and observability.
+
+### Architecture
+
+![OpenCane Architecture](docs/assets/opencane-arch.png)
+
+### Core Capabilities
+
+- Multi-adapter hardware ingress: `mock / websocket / ec600 / generic_mqtt`
+- Multi-modem support in `generic_mqtt`: `ec600mcnle_v1 / a7670c_v1 / sim7600g_h_v1 / ec800m_v1 / ml307r_dl_v1`
+- Realtime voice pipeline: segmentation, VAD, transcription, playback, and interruption
+- Visual lifelog: async image ingestion, semantic retrieval, timeline retrieval, and safety annotations
+- Digital task execution: create, query, cancel, offline pushback, and retry
+- Control APIs: device registration/binding, operation dispatch, and runtime status queries
+
+### Technical Highlights
+
+- Layered architecture: `adapter / runtime / agent / api / storage / safety`
+- Tool execution strategy: MCP-first with automatic fallback to `web_search / web_fetch / exec`
+- Task lifecycle: `pending -> running -> success/failed/timeout/canceled`
+- Persistent data layer: SQLite for lifelog, tasks, and observability
+- Vector retrieval backends: `chroma` and `qdrant`
+- Model provider abstraction via LiteLLM: `OpenAI / Anthropic / Gemini / DashScope` and more
+- Runtime observability with realtime metrics and historical samples
+- Config governance with built-in `dev/staging/prod` profiles and `config check --strict`
+
+### Quick Start
+
+```bash
+git clone https://github.com/iflabx/opencane.git
+cd opencane
+pip install -e .
+```
+
+First-time initialization:
 
-> CLI 主命令为 `opencane`，并兼容 `nanobot` 别名  
-> 项目品牌与产品文档已切换为 OpenCane
+```bash
+opencane onboard
+```
 
-## 架构图
+Apply a profile (staging recommended first):
+
+```bash
+opencane config profile apply --profile CONFIG_PROFILE_STAGING.json
+opencane config check --strict
+```
+
+Start hardware runtime (example):
+
+```bash
+opencane hardware serve --adapter mock --logs
+```
+
+### Documentation
+
+- Overview: `docs/overview.md`
+- Quickstart: `docs/quickstart.md`
+- Architecture: `docs/architecture.md`
+- Data Flow: `docs/data-flow.md`
+- Hardware Runtime: `docs/hardware-runtime.md`
+- Control API: `docs/api/control.md`
+- Lifelog API: `docs/api/lifelog.md`
+- Digital Task API: `docs/api/digital-task.md`
+- Deployment & Config: `docs/deployment-config.md`
+- Operations Runbook: `docs/operations-runbook.md`
+- Security Baseline: `docs/security.md`
+- Roadmap: `docs/roadmap.md`
+
+### NanoBot Compatibility and Migration
+
+- Primary CLI command is `opencane`, with `nanobot` alias compatibility
+- Project branding and product docs are now OpenCane
+- Python package/import path uses `opencane` (no longer supports `from nanobot...`)
+- Default data directory is `~/.opencane`, with backward-compatible reads from `~/.nanobot`
+- Migration guide: `docs/migration-from-nanobot.md`
+- Legacy nanobot docs are archived at: `local-docs/nanobot-legacy/`
+- Archive README: `local-docs/nanobot-legacy/README.md`
+
+### Acknowledgement
+
+OpenCane is continuously developed based on [HKUDS/nanobot](https://github.com/HKUDS/nanobot).  
+Thanks to HKUDS for the open-source foundation.
+
+---
+
+## 简体中文
+
+语言：简体中文 | [Switch to English](#english)
+
+OpenCane 是一个面向智能盲杖场景的 AI 后端运行时，聚焦“设备接入 -> 实时对话 -> 图像记忆 -> 数字任务 -> 安全与观测”的完整闭环。
+
+### 架构图
 
 ![OpenCane 架构图](docs/assets/opencane-arch.png)
 
-## 核心功能
+### 核心功能
 
 - 多硬件接入：支持 `mock / websocket / ec600 / generic_mqtt` 适配器统一接入
 - 多模组适配：`generic_mqtt` 内置 `ec600mcnle_v1 / a7670c_v1 / sim7600g_h_v1 / ec800m_v1 / ml307r_dl_v1`
@@ -18,7 +111,7 @@ OpenCane 是一个面向智能盲杖场景的 AI 后端运行时，聚焦“设
 - 数字任务执行：任务创建、状态查询、取消、离线回推与重试
 - 控制面 API：设备注册绑定、设备指令下发、运行状态查询
 
-## 技术特性
+### 技术特性
 
 - 分层架构：`adapter / runtime / agent / api / storage / safety`
 - 工具执行策略：优先 MCP 工具，失败自动回退 `web_search / web_fetch / exec`
@@ -29,7 +122,7 @@ OpenCane 是一个面向智能盲杖场景的 AI 后端运行时，聚焦“设
 - 运维可观测：运行时指标 + 历史观测数据，支持问题复盘
 - 配置治理：内置 `dev/staging/prod` 模板与 `config check --strict`
 
-## 快速开始
+### 快速开始
 
 ```bash
 git clone https://github.com/iflabx/opencane.git
@@ -56,7 +149,7 @@ opencane config check --strict
 opencane hardware serve --adapter mock --logs
 ```
 
-## 文档导航
+### 文档导航
 
 - 总览：`docs/overview.md`
 - 快速开始：`docs/quickstart.md`
@@ -70,20 +163,18 @@ opencane hardware serve --adapter mock --logs
 - 运维手册：`docs/operations-runbook.md`
 - 安全基线：`docs/security.md`
 - 路线图：`docs/roadmap.md`
-- 迁移说明：`docs/migration-from-nanobot.md`
 
-## 兼容边界
+### NanoBot 兼容与迁移
 
+- CLI 主命令为 `opencane`，并兼容 `nanobot` 别名
+- 项目品牌与产品文档已切换为 OpenCane
 - Python 包与导入路径：使用 `opencane`（不再支持 `from nanobot...`）
-- CLI：主命令 `opencane`，兼容 `nanobot` 别名
 - 数据目录：默认 `~/.opencane`，兼容读取历史 `~/.nanobot`
+- 迁移说明：`docs/migration-from-nanobot.md`
+- 旧版 nanobot 历史文档已归档到：`local-docs/nanobot-legacy/`
+- 归档说明见：`local-docs/nanobot-legacy/README.md`
 
-## 致谢
+### 致谢
 
 本项目基于 [HKUDS/nanobot](https://github.com/HKUDS/nanobot) 进行持续开发与场景化演进。  
 感谢 HKUDS 团队开源 nanobot，为 OpenCane 的落地提供了坚实基础。
-
-## 历史文档
-
-旧版 nanobot 历史文档已归档到：`local-docs/nanobot-legacy/`  
-归档说明见：`local-docs/nanobot-legacy/README.md`