An AI-powered discovery engine that bypasses commercial platforms to find authentic local recommendations for food, travel, and lifestyle.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install local-guide
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install local-guide using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Local Guide?
The Local Guide skill is a specialized search agent designed to uncover authentic experiences by bypassing traditional commercial rating platforms like Meituan, Ctrip, or Dianping. By utilizing a multi-layered search strategy, it filters through the noise of algorithm-driven rankings and paid promotions to identify hidden gems truly recognized by local residents. Whether searching for a decade-old hole-in-the-wall eatery or a cost-effective parking spot, this Openclaw Skills integration ensures data integrity through anti-spam filtering.
This skill prioritizes official government culture and tourism recommendations, local news archives, and native community discussions over commercial influencers. It provides users with a transparent view of every location, including documented pros and cons, to ensure a reliable and grounded travel or dining experience.
Local Guide Use Cases
- Finding authentic local snacks and time-honored restaurants (Laozihao) in a specific city district.
- Locating cost-effective and convenient parking near busy hospitals or scenic landmarks.
- Discovering niche, non-commercialized tourist spots and weekend activities known only to residents.
- Identifying high-value hotels and hot spring resorts that locals actually frequent.
- Avoiding tourist traps and heavily marketed influencer locations on social media.
How Local Guide Works
- The agent receives a trigger command specifying the category (e.g., food, parking, scenery) and the target location.
- It executes a rigorous Anti-Spam Filter to exclude data from major commercial rating and booking platforms.
- A multi-round deep search (5-8 rounds) is initiated across government archives, local news media, and community forums like Douban or local neighborhood boards.
- Search queries are dynamically generated using specific intent keywords like "only locals know" or "government recommended."
- Data is cross-referenced across multiple independent sources to verify authenticity.
- Findings are structured into a detailed report containing basic info, "authentic evidence" reasons, and real-world warnings (disadvantages).
Local Guide Setup
To integrate this capability into your AI agent or send results to Feishu, ensure you have the Openclaw Skills environment configured. For Feishu card integration, use the following command structure:
# Example: Sending discovered local data to Feishu
echo '{"location": "Guangzhou", "type": "Food", "items": [...]}' | python3 scripts/send_feishu_card.py {receive_id} {reply_to}
Local Guide Data Schema & Taxonomy
The skill organizes output into two primary schemas depending on the query type:
General Recommendations (Food, Hotels, Spots)
| Field | Description |
|---|---|
| Name & Location | Specific street or landmark reference |
| Navigation | Direct AMAP (Gaode) compatible address |
| Price Range | Per capita consumption (e.g., ¥30-50) |
| Authentic Evidence | Why it is considered local (e.g., 10+ years in business) |
| Reality Check | Cons/Warnings (e.g., cash only, poor service) |
Parking Specific Schema
| Field | Description |
|---|---|
| hour_rate | Cost per hour (Mandatory) |
| spaces | Total estimated parking spots |
| daily_cap | Maximum daily charge limit |
| charging | Presence of EV charging piles |
Local Guide Advanced Features
- Multi-round recursive search (up to 8 rounds) to ensure at least 10 high-quality results per category.
- Platform-aware formatting: Automatically generates rich Feishu cards when a Feishu channel ID is detected.
- Negative feedback integration: Explicitly searches for and includes "Reality Checks" (disadvantages) to combat marketing bias.
- Source transparency: Lists the specific non-commercial sources used for the final recommendation set.
SKILL.md
Loading
Related Openclaw Skills
Happy Comedy
A creative AI skill that transforms random entries from the HappyDB dataset into punchy, observational stand-up comedy routines.
Openclaw Dynamics 365 CRM Skill
A powerful integration for AI agents to create and manage Microsoft Dynamics 365 CRM records using the Dataverse Web API.
CementOps Environmental Compliance Advisor
An expert AI advisor for cement plant environmental managers to navigate EPA regulations, NESHAP limits, and CEMS QA/QC workflows.
ADHD Daily Planner
A specialized planning system designed to work with the ADHD brain using dopamine-aware strategies and realistic time estimation.
AdMapix Intelligence Assistant
An advanced AI assistant for ad intelligence and mobile app analytics powered by the AdMapix API for data-driven market insights.
Video Editor CapCut
A powerful AI-driven video editing skill that automates trimming, text overlays, and music syncing for short-form content.
Featured*
