daolv-hotel-search

Original source / Raw SKILL.md

---
name: daolv-hotel-search
description: Hotel search and shortlist comparison using the ai-go-hotel MCP server (getHotelSearchTags, searchHotels). Use when users need fast hotel discovery by city/POI/date/budget/amenities/brand, but do not yet need full booking-room confirmation.
---

# Daolv Hotel Search

Focus on fast discovery and clean shortlist output.

## Workflow

1. Capture search intent
- Extract destination, check-in date, nights, guests, budget, star range, distance preference, amenities, brand preferences.
- Ask only minimal missing questions.

2. Prime tags once
- Call `ai-go-hotel.getHotelSearchTags` once per task.
- Map user intent into `hotelTags.requiredTags` / `preferredTags` / `excludedTags`.

3. Run hotel search
- Call `ai-go-hotel.searchHotels` with required fields:
  - `originQuery`, `place`, `placeType`
- Include optional fields when available:
  - `checkInDate`, `stayNights`, `adultCount`, `countryCode`, `starRatings`, `distanceInMeter`, `size`, `hotelTags`
- Prefer `size=10` first, then narrow to top 3-5.

4. Return shortlist only
- Output: 1 recommended + 2 alternatives.
- Include quick trade-off: price / location / amenities / star.
- Do not deep-dive room plans unless user asks to proceed to booking.

## Output Template

- **搜索条件**：地点 / 日期 / 人数 / 预算 / 偏好
- **首选酒店**：价格区间、位置、标签命中、推荐理由
- **备选 1-2**：同结构
- **下一步**：是否进入“房型与退改细节”阶段

## Quality Bar

- Use concrete numbers when available.
- Use `price.lowestPrice` + `price.currency` (price is object).
- If price missing/null, mark as “需二次查价”.
- Keep output concise and decision-oriented.

## MCP Preset Config

- Embedded preset: `references/mcp-client-config.json`
- Endpoint: `https://mcp.aigohotel.com/mcp` (`streamable_http` + prefilled Authorization header)


---

## Referenced Files

> The following files are referenced in this skill and included for context.

### references/mcp-client-config.json

```json
{
  "mcpServers": {
    "aigohotel-mcp": {
      "url": "https://mcp.aigohotel.com/mcp",
      "type": "streamable_http",
      "headers": {
        "Authorization": "Bearer mcp_171e1ffa7da343faa4ec43460c52b13f",
        "Content-Type": "application/json"
      }
    }
  }
}

```



---

## Skill Companion Files

> Additional files collected from the skill directory layout.

### _meta.json

```json
{
  "owner": "cnchenkai",
  "slug": "daolv-hotel-search",
  "displayName": "Daolv Hotel Search",
  "latest": {
    "version": "0.1.0",
    "publishedAt": 1771992312801,
    "commit": "https://github.com/openclaw/skills/commit/3dd1edc004b754c654b91ffb7846886a629eda47"
  },
  "history": []
}

```

### references/aigohotel-mcp-spec.md

```markdown
# AIGoHotel MCP Live Spec (2026-02-11)

Source snapshot provided by user.

## Tools

1. `searchHotels`
2. `getHotelDetail`
3. `getHotelSearchTags`

## Key Compatibility Notes

- `searchHotels.price` is an object (not a scalar number).
  - Prefer: `price.lowestPrice`, `price.currency`, `price.hasPrice`.
- `searchHotels` may return nullable or missing fields by supplier.
- `getHotelDetail` may fail with plain text error (e.g. “获取价格失败，请稍后重试”).
- `roomRatePlans` can be large; clients should cap displayed rows.
- Date defaults/fallbacks are server-side tolerant:
  - `checkInDate` invalid/past/empty → tomorrow
  - `checkOutDate` invalid/non-later → `checkInDate + 1 day`

## searchHotels required fields

- `originQuery` (string)
- `place` (string)
- `placeType` (string)

`placeType` supports:
- 城市, 机场, 景点, 火车站, 地铁站, 酒店, 区/县, 详细地址

## searchHotels optional highlights

- `queryParsing` (default true)
- `checkInDate` (`YYYY-MM-DD`)
- `stayNights` (default 1)
- `adultCount` (default 2)
- `countryCode` (ISO 3166-1 alpha-2)
- `distanceInMeter` (POI scenario)
- `starRatings` (default `[0.0,5.0]`, step 0.5)
- `size` (default 5, max 20)
- `withHotelAmenities`
- `language` (default `zh_CN`)
- `hotelTags`
  - `preferredTags`
  - `requiredTags`
  - `excludedTags`
  - `preferredBrands`
  - `maxPricePerNight`
  - `minRoomSize`

## getHotelDetail input highlights

- one of: `hotelId` or `name` (hotelId wins)
- `checkInDate`, `checkOutDate`
- `adultCount`, `childCount`, `childAgeDetails`
- `roomCount`
- `countryCode` (default CN)
- `currency` (default CNY)

## getHotelSearchTags output use

- fetch once and cache per task
- map intent to:
  - `hotelTags.preferredTags`
  - `hotelTags.requiredTags`
  - `hotelTags.excludedTags`

## Embedded Client Preset

- See `references/mcp-client-config.json` for the prefilled server config and Authorization header.

## Links

- Source repo: https://github.com/longcreat/aigohotel-mcp
- API key apply: https://mcp.agentichotel.cn/apply
- MCP spec: https://modelcontextprotocol.io/

```

### references/distribution.md

```markdown
# Distribution Checklist (ClawHub + awesome-openclaw-skills)

## 1) Preflight

- Ensure skill folder name equals frontmatter `name`.
- Validate frontmatter has only `name` and `description`.
- Remove placeholders/TODO text.
- Run package command:

```bash
python3 /usr/lib/node_modules/openclaw/skills/skill-creator/scripts/package_skill.py \
  /root/.openclaw/workspace/skills/daolv-hotel-booking \
  /root/.openclaw/workspace/dist
```

## 2) Publish to ClawHub

```bash
clawhub login
clawhub publish /root/.openclaw/workspace/skills/daolv-hotel-booking
```

If not logged in, complete browser login first, then rerun publish.

## 3) Publish to awesome-openclaw-skills (GitHub)

1. Fork `awesome-openclaw-skills`.
2. Add one concise entry under the relevant category (travel/hospitality if present).
3. Include:
   - Skill name + 1-line value proposition
   - ClawHub link (or repo link if pending)
4. Open PR with title:
   - `Add daolv-hotel-booking skill`

Recommended commit message:
- `feat: add daolv-hotel-booking to travel skills`

## 4) Post-publish verification

- Install from registry in a clean path and verify:

```bash
mkdir -p /tmp/skill-smoke && cd /tmp/skill-smoke
clawhub install <published-slug>
```

- Run one smoke query:
  - “帮我找下周武汉出差酒店，预算 500/晚，靠近地铁，含早餐。”

## 5) Changelog discipline

For every update, note:
- What changed (workflow, parameters, output format)
- Why it changed (bugfix, better ranking, better UX)
- Compatibility impact (if any)

```

### references/promo-copy.md

```markdown
# Promo Copy Templates

## A) ClawHub short description

面向酒店预订与旅行规划的实战 Skill：基于道旅 MCP 自动完成“需求澄清 → 酒店筛选 → 房型比价 → 决策建议 → 下单前确认”，输出可直接执行的预订方案。

## B) Social / group announcement

🚀 新技能上线：`daolv-hotel-booking`

如果你在 OpenClaw 里经常遇到“找酒店、比价格、选房型、做决策”这类任务，这个 Skill 可以直接用：

- 基于道旅 MCP（酒店搜索 + 详情）
- 自动提取预算/位置/人群/设施偏好
- 输出“首选 + 两个备选”并附带取舍理由
- 给出最终下单前 2-4 条关键确认项

适用场景：出差、亲子游、自由行、临时改签住宿。

## C) awesome-openclaw-skills PR body

### What this adds

Add `daolv-hotel-booking`, a travel/hotel planning skill powered by ai-go-hotel MCP.

### Why it is useful

- Turns ambiguous booking requests into structured constraints
- Produces shortlist + room-level detail comparison
- Reduces decision latency with explicit trade-off guidance

### Typical user request

“下周去上海出差，预算每晚 700，地铁步行 10 分钟内，帮我选 3 家并说明推荐理由。”

```