ai-assistant-skills

What This Skill Does

AI-Assistant-Skills is a powerful analytical engine designed for automotive sales professionals. It acts as an intelligent layer between raw sales transcripts and actionable business intelligence. The skill parses unstructured dialogue, filters out noise, and reconstructs the interaction into a clean, professional-grade summary. By focusing on explicit and implicit customer needs, it allows sales teams to bypass tedious note-taking and jump straight into high-conversion follow-up strategies. It extracts critical data points across five key dimensions—car model preference, budget, usage scenarios, decision blockers, and store visit status—ensuring no lead goes cold.

Installation

To add this skill to your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/dadaliu0121/ai-assistant-skills

Use Cases

This skill is designed for high-volume sales environments where rapid analysis leads to faster deal closures. Use it during daily pipeline reviews to identify which leads require immediate attention, or after a long day of showroom walk-ins to organize your CRM logs. It is particularly effective when you have lengthy chat logs or transcriptions that contain too much social fluff and not enough actionable customer intent.

Example Prompts

1. "分析这段销售对话: [Paste Transcript Here]"
2. "请帮我总结一下这位客户的购车意向，重点分析他的决策阻碍在哪里。"
3. "我已经和客户沟通了半小时，但我感觉他还在犹豫，你能给出具体的销售跟进建议吗？"

Tips & Limitations

• Quality of Input: The quality of the analysis is directly proportional to the detail provided in the conversation transcript. Ensure your transcripts include both the sales rep's questions and the customer's raw responses.
• Implicit Detection: The AI excels at identifying subtle hints (e.g., mentioning fuel prices implies budget sensitivity), but it will never invent facts. If the information isn't present, it will intelligently mark it as [需跟进] (needs follow-up).
• Conciseness: To get the best results, aim for transcripts that highlight the core of the deal. Keep the analysis focused by strictly adhering to the three-part output structure requested by the system. Avoid uploading highly non-related small talk to save processing time.