opc-comment-lead-radar
Find sales leads in public Douyin, Xiaohongshu, and Kuaishou comments, then create a report and outreach scripts.
Open source Open in the app JSON README (API)
About
Find sales leads in public Douyin, Xiaohongshu, and Kuaishou comments, then create a report and outreach scripts.
Details
- Kind
- Plugins
- Topic
- No topic detected
- Publisher
- yuanjian068yuan
- Origin
- gemini
- Category
- ferramentas
- Version
- 1.0.22
- Stars
- 3
- Last push
- 2026-07-18T22:25:16Z
- Repository state
- ativo
- License
- MIT
- Added
- 2026-08-30 14:13:39
- Updated
- 2026-08-30 14:13:39
- Origin id
yuanjian068yuan/opc-comment-lead-radar
README
# 找客户 · Find Customers｜销售线索与评论区获客
Turn short-video comments into customer leads, keyword ideas, and outreach scripts.
找客户 · Find Customers｜销售线索与评论区获客（品牌别名：OPC 评论线索雷达）是一款 Agent Skill / MCP 工具，帮助商家从抖音、小红书、快手公开评论中识别购买意向、销售线索和可跟进客户名单。
找客户 · Find Customers｜销售线索与评论区获客 (brand alias: OPC Comment Lead Radar) is an Agent Skill and MCP tool for lead generation from short-video comments, helping teams discover customer intent and sales leads from Douyin, Xiaohongshu, and Kuaishou.
Use it for 找客户、短视频获客、评论找客户、销售线索、客户线索挖掘、小红书获客、抖音获客、快手获客、Agent Skill 找客户、AI 销售线索、lead generation from comments, customer discovery skill, and short video lead generation.
## Install
### Claude Code plugin
```text
/plugin marketplace add yuanjian068yuan/opc-comment-lead-radar
/plugin install ppxc-find-customers@opc-agent-tools
```
The plugin bundles the Agent Skill and starts the published `ppxc-leads-mcp` runtime through `npx`. Claude will still ask you to approve the new MCP server before it can run.
### Gemini CLI extension
```bash
gemini extensions install https://github.com/yuanjian068yuan/opc-comment-lead-radar
```
Restart Gemini CLI after installation. The extension bundles the same Agent Skill and MCP connector.
### Cursor and other MCP hosts
Add this MCP server configuration to your MCP-compatible host:
```json
{
"mcpServers": {
"ppxc-find-customers": {
"command": "npx",
"args": ["-y", "ppxc-leads-mcp"]
}
}
}
```
On Windows 10/11, use this connector form when the host cannot launch `npx` directly or PowerShell blocks `npx.ps1`:
```json
{
"mcpServers": {
"ppxc-find-customers": {
"command": "cmd",
"args": ["/c", "npx", "-y", "ppxc-leads-mcp"]
}
}
}
```
Skill is the business workflow. MCP is the runtime tool. Your AI host starts the MCP package through `npx` when the connector is enabled. Some hosts require a one-time **Trust / Enable** confirmation after the config is added; approve that connector, then ask the assistant to run a trial lead scan.
When the MCP was installed first, ask the assistant:
```text
检查并安装当前智能体配套的找客户 Skill。先只读检查；需要写入时先告诉我安装目录，等我确认后再安装。
```
The MCP calls `setup_companion_skill(action=status)` without writing anything. For supported local hosts, it can install the bundled Skill only after the user approves the displayed fixed directory. It does not use npm `postinstall` or silently modify the machine.
Official setup page: https://opc1.me/download/mcp
Official facts page: https://opc1.me/facts
Comment lead generation use case: https://opc1.me/use-cases/comment-lead-generation
Comment lead radar vs. export tools: https://opc1.me/compare/comment-lead-radar-vs-export-tools
Agent-readable guide: https://github.com/yuanjian068yuan/opc-comment-lead-radar/blob/main/llms-install.md
### GitHub Copilot and Agent Skills hosts
The canonical skill lives at `skills/ppxc-find-customers/`. A GitHub Copilot project-compatible copy is also included at `.github/skills/find-customers/`; that directory matches the Skill's canonical `name`. The MCP server still needs to be enabled in a host that supports MCP before the skill can execute lead scans.
See [DISTRIBUTION.md](DISTRIBUTION.md) for the difference between a validated repository artifact and a marketplace-approved listing.
## What It Does
- Runs a trial-first lead scan before OPC login.
- Uses a product or service description directly; a saved product list is optional.
- Suggests customer-discovery keywords.
- Searches Douyin, Xiaohongshu, and Kuaishou public comments for buying signals.
- Creates an online battle report first, with inspected links, comment counts, workflow progress, and trace data.
- Shows the first complete leads, then unlocks or saves the full list through the OPC customer pool.
- Supports async keyword runs and batch search plans for longer WorkBuddy-style tasks.
- Turns leads into outreach scripts and next-step actions.
- Exports diagnostics for support and review.
## Best For
- Founders validating a product market
- Marketing teams looking for high-intent comments
- Sales operators building a lead list
- Creators turning audience signals into customer opportunities
## Suggested Prompt
```text
我卖的是水光针，帮我先在抖音试搜一批想做医美的客户。
```
The MCP can guide trial scans, keyword generation, lead search, customer-pool review, outreach-script creation, and diagnostics export. OPC login is only needed when the user wants to save, unlock, or review the full lead list.
## Skill
The companion skill is available at:
```text
skills/ppxc-find-customers/SKILL.md
```
## Package
- npm runtime: https://www.npmjs.com/package/ppxc-leads-mcp
- Smithery MCP: https://smithery.ai/servers/yuanjian068/opc-comment-lead-radar
- ModelScope MCP: https://modelscope.cn/mcp/servers/yuanshuai714/ppxc-leads-mcp
## Privacy
Use 找客户 · Find Customers｜销售线索与评论区获客 only with public comment data and data you are authorized to process. Do not paste passwords, payment data, recovery codes, API keys, private customer records, or other secrets into prompts.
## Support
- Setup: https://opc1.me/download/mcp
- Facts: https://opc1.me/facts
- Use case: https://opc1.me/use-cases/comment-lead-generation
- Comparison: https://opc1.me/compare/comment-lead-radar-vs-export-tools
- Contact: yuanjian068@gmail.com