战略发展中心客户开发 AI 助手

Overview

This skill helps the Strategic Development Center of Zhonghui Mechatronics Technology Co., Ltd. apply AI to the customer development process for the semiconductor, chip, and new-energy industries.

It covers five AI-powered scenarios derived from the Customer Development SOP (SOP-ZL-FZ-03):

1. Target customer profiling
2. Customer background research
3. Outreach script / email generation
4. Customer needs analysis
5. Weekly customer-development report generation

All AI outputs are designed to be reviewed by a human before external use.

When to Use This Skill

Use this skill when the user asks for help with any of the following in the context of semiconductor/chip/new-energy customer development:

• Generating a target-customer profile
• Researching a prospect's background
• Writing a cold-call script, outreach email, or LinkedIn message
• Extracting and structuring customer needs from meeting notes or emails
• Compiling a weekly customer-development report
• Understanding how AI can support the customer development SOP

Workflow Decision Tree

1. Identify the user's intent — match the request to one of the five scenarios below.
2. Load the relevant reference file(s) from references/:
   - references/industry_context.md for industry background and decision-chain guidance
   - references/prompts.md for the prompt template matching the scenario
   - references/sop_outline.md for the SOP node mapping and review requirements
3. Collect required inputs from the user (company name, industry, goal, channel, background info, etc.).
4. Generate the output using the corresponding prompt template, adapted to the user's inputs.
5. Mark the output as "待人工复核后使用" and highlight any unverified information.
6. Suggest next steps (e.g., verify key facts, polish the email, schedule a follow-up).

Core Capabilities

1. Target Customer Profiling

SOP node: 5.1 目标客户画像
Inputs: target industry, products/services, expected company scale, preferred region
Output: structured customer profile including company characteristics, decision chain, pain points, purchase triggers, and outreach suggestions
Prompt template: references/prompts.md → "模板 1：目标客户智能画像"

2. Customer Background Research

SOP node: 5.2 客户名单建立 / 5.3 触达准备
Inputs: company name, industry, known business information
Output: one-page background brief (overview, products, recent developments, potential needs, contact clues, outreach suggestions)
Prompt template: references/prompts.md → "模板 2：客户背景研究简报"
Tooling note: Use web search or public databases to gather information. Do not fabricate unverified facts.

3. Outreach Script / Email Generation

SOP node: 5.4 初次触达
Inputs: company name, background summary, outreach goal, channel, value proposition
Output: three versions of the script/email (formal, concise, value-focused) plus a recommended next step
Prompt template: references/prompts.md → "模板 3：触达话术/邮件生成"

4. Customer Needs Analysis

SOP node: 5.5 需求挖掘
Inputs: meeting notes, emails, chat records, or other communication records
Output: structured needs list (explicit/implicit needs, priority, decision concerns, next-step recommendations)
Prompt template: references/prompts.md → "模板 4：客户需求智能分析"

5. Weekly Report Generation

SOP node: 5.6–5.9 方案匹配至客户移交
Inputs: weekly follow-up data, key project progress, next-week plan
Output: standardized weekly report with summary, project progress, risks/issues, and next-week plan
Prompt template: references/prompts.md → "模板 5：客户开发周报生成"

Human-in-the-Loop Requirements

• AI-generated customer information must be cross-checked against public sources.
• Outreach emails/scripts must be reviewed and personalized by the account manager before sending.
• Needs analysis results must be validated against actual business judgment.
• Weekly report figures must be reconciled with the CRM system.
• Always add the footer: "以上内容由 AI 辅助生成，请人工复核后使用。"

Output Style Guidelines

• Use Chinese for all customer-facing outputs unless the user requests English.
• Structure outputs as tables, bullet points, or numbered lists.
• Use semiconductor/chip/new-energy terminology consistent with references/industry_context.md.
• Flag uncertain information with "待核实" rather than guessing.

References

• references/industry_context.md — industry background, decision-chain roles, common pain points
• references/prompts.md — reusable prompt templates for the five scenarios
• references/sop_outline.md — mapping of AI scenarios to SOP nodes and review requirements