index1-doctor

TotalClaw 作者 gladego v2.0.3

诊断index1环境 - 检查Python、Ollama、模型、索引运行状况。

安装 / 下载方式

TotalClaw CLI推荐
totalclaw skills install shop:totalclaw:totalclaw~gladego-index1-doctor
cURL直接下载,无需登录
curl -fsSL https://skills.taituai.com/api/skills/totalclaw%3Atotalclaw~gladego-index1-doctor/file -o gladego-index1-doctor.md
# index1 Doctor

Environment diagnostic skill for index1. Runs health checks and provides fix recommendations.

## Usage

Type `/doctor` or ask the agent to diagnose index1.

## What it checks

The skill runs three commands sequentially and analyzes results:

### 1. Environment Check

```bash
index1 doctor
```

Checks:
- Python version (>= 3.10 required)
- SQLite version (>= 3.43.0 for full features)
- sqlite-vec extension
- ONNX embedding (built-in, bge-small-en-v1.5)
- Ollama connectivity (optional, for multilingual/CJK)
- Embedding model availability
- CJK/Chinese support (jieba)

### 2. Index Status

```bash
index1 status
```

Shows:
- Document count and chunk count
- Collections list
- Last index time
- Database size

### 3. Ollama Models

```bash
ollama list
```

Shows installed models. Recommended embedding models:
- `nomic-embed-text` — Standard, 270MB
- `bge-m3` — Best for CJK content, 1.2GB

## Interpreting Results

| Check | Pass | Fail Fix |
|-------|------|----------|
| Python | >= 3.10 | Install Python 3.11+ |
| SQLite | >= 3.43.0 | Auto-degrades, no action needed |
| sqlite-vec | Loaded | `pip install index1` (bundled) |
| Ollama | Connected | `curl -fsSL https://ollama.com/install.sh \| sh` |
| Model | Available | `ollama pull nomic-embed-text` |
| CJK | jieba loaded | `pip install index1[chinese]` |
| Index | Has documents | `index1 index ./src ./docs` |

## When to use

- First-time setup verification
- After upgrading index1
- When search returns unexpected results
- When vector search stops working
- Before reporting issues

---

## 中文说明

# index1 Doctor

index1 的环境诊断 skill。运行健康检查并提供修复建议。

## 用法

输入 `/doctor` 或要求代理诊断 index1。

## 它检查什么

该 skill 依次运行三条命令并分析结果:

### 1. 环境检查

```bash
index1 doctor
```

检查项:
- Python 版本(需 >= 3.10)
- SQLite 版本(完整功能需 >= 3.43.0)
- sqlite-vec 扩展
- ONNX embedding(内置,bge-small-en-v1.5)
- Ollama 连通性(可选,用于多语言/CJK)
- Embedding 模型可用性
- CJK/中文支持(jieba)

### 2. 索引状态

```bash
index1 status
```

显示:
- 文档数量和分块数量
- 集合列表
- 上次索引时间
- 数据库大小

### 3. Ollama 模型

```bash
ollama list
```

显示已安装的模型。推荐的 embedding 模型:
- `nomic-embed-text` —— 标准款,270MB
- `bge-m3` —— 最适合 CJK 内容,1.2GB

## 解读结果

| 检查项 | 通过 | 失败修复 |
|-------|------|----------|
| Python | >= 3.10 | 安装 Python 3.11+ |
| SQLite | >= 3.43.0 | 自动降级,无需操作 |
| sqlite-vec | 已加载 | `pip install index1`(已捆绑) |
| Ollama | 已连接 | `curl -fsSL https://ollama.com/install.sh \| sh` |
| Model | 可用 | `ollama pull nomic-embed-text` |
| CJK | jieba 已加载 | `pip install index1[chinese]` |
| Index | 有文档 | `index1 index ./src ./docs` |

## 何时使用

- 首次设置验证
- 升级 index1 之后
- 当搜索返回意外结果时
- 当向量搜索停止工作时
- 报告问题之前