WeChat Contact List Video Extraction & CRM Analyzer
Zero-risk, permission-free WeChat personal contacts extraction and structured CRM pipeline. Uses macOS native screen recording (Cmd+Shift+5), high-FPS FFmpeg frame extraction, Apple Silicon concurrent Vision OCR, and multi-dimensional rule-based NLP extraction (Name, Title, Org, Venue, City, Time) to generate Excel/CSV, JSON, and interactive HTML dashboards. Skill: WeChat Contact List Video Extraction & CRM Analyzer Owner: emergencescience Summary: Zero-risk, permission-free WeChat personal contacts extraction and structured CRM pipeline. Uses macOS native screen recording (Cmd+Shift+5), high-FPS FFmpeg frame extraction, Apple Silicon concurrent Vision OCR, and multi-dimensional rule-based NLP extraction (Name, Title, Org, Venue, City, Time) to generate Excel/CSV, JSON,
Rank
62
Safety
84
Downloads
1.4k
Updated
Oct 10, 2026
Version
0.1.0
Source
CLAWHUB
About
What it does, and when to use it.
Capability contract not published. No trust telemetry is available yet. 1.4K downloads reported by the source. Last updated 10/10/2026.
Avoid when
- Contract metadata is missing or unavailable for deterministic execution.
Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing
Public facts
Every fact links back to the source it came from.
- Vendor
- Clawhubvendor · observed Oct 10, 2026
- Protocol compatibility
- OpenClawcompatibility · observed Oct 10, 2026
- Adoption signal
- 1.4K downloadsadoption · observed Oct 10, 2026
- Latest release
- 0.1.0release · observed Aug 17, 2026
- Handshake status
- UNKNOWNsecurity
Install and run
Setup complexity: low.
clawhub skill install s17a35617mabw6kewbzv4z4e7x83w38h:emergence-wechat-contact-crm- Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.
- Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.
Contract: missing
curl -s "https://www.xpersona.co/api/v1/agents/clawhub-emergencescience-emergence-wechat-contact-crm/snapshot"
Documentation
CLAWHUB
12,003 characters of source documentation, loaded on request.
Extracted files
5 files captured from the source.
SKILL.md
---
slug: emergence-wechat-contact-crm
name: WeChat Contact List Video Extraction & CRM Analyzer
version: 1.0.0
description: Zero-risk, permission-free WeChat personal contacts extraction and structured CRM pipeline. Uses macOS native screen recording (Cmd+Shift+5), high-FPS FFmpeg frame extraction, Apple Silicon concurrent Vision OCR, and multi-dimensional rule-based NLP extraction (Name, Title, Org, Venue, City, Time) to generate Excel/CSV, JSON, and interactive HTML dashboards.
---
# 微信通讯录视频录屏提取与智能 CRM 看板构建技能
本技能提供了一套**完全无需系统辅助功能权限、零封号风险、本地纯离线**的个人微信海量通讯录提取与资产化方案。
通过 macOS 自带录屏快捷键(`Cmd + Shift + 5`)录制联系人列表平稳滑动视频,结合高帧率抽帧、Apple Vision 硬件加速 OCR 与多维实体解析,可在 30 秒内完整提取 5000+ 联系人并生成交互式 CRM 看板。
---
## 1. 核心架构与工作流
```mermaid
graph TD
A[MacBook Cmd+Shift+5 录屏] -->|生成 .mov 视频| B[FFmpeg 40 FPS 高频抽帧]
B --> C[macOS 原生 Apple Vision OCR]
C -->|8线程并发 毫秒级识别| D[时序去重 & Overlap 统计引擎]
D -->|输出 raw 纯文本| E[结构化规则/大模型 NLP 抽取]
E --> F1[Excel / CSV 导出]
E --> F2[标准 JSON 数据库]
E --> F3[现代化 HTML 资产看板]
```
---
## 2. 命名提取模型与打标习惯建议 (Naming Schema & Best Practices)
本技能的结构化深度依赖于用户日常对微信好友的打标与备注习惯。推荐的最佳实践范式为:
$$\text{好友备注} = [\text{微信名/称谓}] + [\text{职位/头衔/机构/学校}] + [\text{场景/场地/活动/商圈}] + [\text{时间/年份}]$$
### 提取维度:
- **微信名 / 核心称呼**:如 `张三`, `李四`, `Alex`。
- **职位 / 头衔 / 机构**:如 `合伙人`, `架构师`, `某某大学`, `某某科技` 等。
- **场景 / 场地 / 圈子**:如 `某某峰会`, `某某沙龙`, `某某商圈`, `某某酒店`。
- **城市 / 地区**:支持一线城市、主要省会与全球核心枢纽。
- **时间 / 年份**:支持 4 位年份(如 `2023`, `2024`)。
- **行业归类**:自动映射至 `金融/投资`, `科技/互联网/AI`, `法律/法务/咨询`, `高校/科研/教育`, `医疗/健康`, `文化/传媒/消费` 等。
> 💡 **给用户的打标建议**:
> 微信本身缺乏多维标签系统,日常在备注中顺手记录「认识地点与年份」,不仅便于日常搜索唤醒人脉,也能在批量建档时最大化发挥数据资产价值。
---
## 3. 实测数据与录屏滑动速度指南
在 5000+ 联系人库上的真实基准测试数据:
| 录屏时长 | 提取帧数 (@40-50 FPS) | 捕获联系人数 | 覆盖率 | 相邻帧平均重叠人数 | 适用场景 |
| :--- | :--- | :--- | :--- | :--- | :--- |
| **5 秒快速录制** | 248 帧 | 1,607 人 | ~30% | 3.2 人 | 快速抽样/测试连通性 |
| **20 秒平稳录制** | 798 帧 | 4,389 人 | **81.7%** | 4.1 人 | 日常快速全量备份 |
| **30~40 秒匀速录制** (推荐) | ~1400 帧 | 5,000+ 人 | **>95%** | 8~12 人 | **全量无损建档 (推荐)** |
> 💡 **最佳录制建议**:
> 用双指在 Mac 触控板上从 A 到 Z 匀速轻推下滑,耗时约 **30 秒** 即可达到极高重叠率(>60%),做到 0 跳帧、无死角覆盖。
---
## 4. 命令行执行指引
### 步骤 1:录制视频
按下 `Command + Shift + 5`,框选桌面微信联系人列表区域,点击录制,匀速下滑至底部后停止,保存为 `wechat_contacts.mov`。
### 步骤 2:从视频提取联系人
```bash
python scripts/process_video.py wechat_contacts.mov --output data/contacts_raw.txt --fps 40.0
```
### 步骤 3:一键结构化并生成看板
```bash
python scripts/structure_contacts.py data/contacts_raw.txt --csv data/contacts.csv --html data/contacts_crm.html
```
*(可选) 如需使用自定义场景词表或企业词表:*
```bash
python scripts/structure_contacts.py data/contacts_raw.txt --config config.example.json --html data/contacts_crm.html
```
---
## 5. 依赖环境
- **macOS**:macOS 12.0+ (原生内置 Vision.framework,无需额外安装 OCR 库)
- **Swift 编译器**:macOS 自带 `swiftc`(Xcode Command Line Tools)
- **FFmpeg**:`brew install ffmpeg`
- **Python**:Python 3.9+_meta.json
{
"ownerId": "kn7c83ffhzqbhaqes1fdt9t5e5825jme",
"slug": "emergence-wechat-contact-crm",
"version": "0.1.0",
"publishedAt": 1786968214881
}README_zh.md
# 微信人脉资产智能 CRM 提取技能 (WeChat Contacts Video to CRM)
> 🚀 **零系统权限困扰 · 零微信封号风险 · 纯本地硬件加速 · 5000+ 联系人 30 秒一键结构化与可视化建档**
---
## 🌟 为什么开发这个技能?
对于微信好友数高达 3000~5000+ 的创业者、商务拓展(BD)、投资人、社群主理人和开发者来说,微信通讯录是极其庞大的人脉资产。但个人号面临两大痛点:
1. **官方无开放 API**:个人微信没有合法的外部导出接口。
2. **安全风控极高**:使用内存 Hook、注入 DLL 或协议挂机类外挂软件极易触发封号,且 macOS 的 SIP 和辅助功能权限机制经常拦截模拟点击脚本。
本项目通过 **「物理黑盒无感录屏 + macOS 原生 Apple Vision 毫秒级 OCR + 多维时空 NLP 结构化」**,彻底解决了上述所有痛点!
---
## ⚡ 核心实测数据:你应该滑多快?(录屏速度指南)
我们在 **5000+ 位** 真实微信好友的大盘上进行了全量实测与相邻帧重叠率(Overlap Ratio)统计:
```
+----------------------------------------------------------------------------------------------------+
| 录屏时长 | 采样帧率 (@FPS) | 提取总帧数 | 捕获联系人数量 | 覆盖率 | 相邻帧平均重叠人数 | 推荐度 / 场景 |
+----------------------------------------------------------------------------------------------------+
| 5 秒 | 50 FPS | 248 帧 | 1,607 人 | ~30% | 3.2 人 / 帧 | ⚠️ 过快 (有跳帧) |
| 20 秒 | 40 FPS | 798 帧 | 4,389 人 | 81.7% | 4.1 人 / 帧 | ⚡ 快速全量备份 |
| 30~40 秒 | 40 FPS | ~1400 帧 | 5,000+ 人 | >95% | 8~12 人 / 帧 | 🌟 推荐 (0跳帧无损) |
+----------------------------------------------------------------------------------------------------+
```
### 💡 黄金录制法则:
- **单屏可见约 10~14 人**。
- 如果录制时间为 **30 秒**,以 40 FPS 抽帧计算,相邻两帧之间平均有 **8~12 人重叠**(重叠率 > 70%)。
- 这种平稳重叠不仅能 **100% 消除滑动动态模糊**,还能实现 **0 跳帧漏抓**。
---
## 🎬 3 步操作上手指南
### 1. 录制联系人列表视频
1. 打开 Mac 桌面端微信,切换到「通讯录」选项卡,列表滑动到最顶部。
2. 按下 Mac 原生录屏快捷键 **`Command + Shift + 5`**。
3. 选择 **「录制所选部分」**,用选框精准框住联系人列表区域。
4. 点击「录制」,双指在触控板上从 A 到 Z 匀速下滑(耗时约 25~35 秒),滑到底部后点击顶部菜单栏停止录屏,文件将保存为 `.mov`。
### 2. 从视频提取联系人
```bash
# 运行视频抽帧与原生 OCR 引擎
python scripts/process_video.py wechat-contacts.mov --output data/contacts_raw.txt --fps 40.0
```
> *注:脚本会自动调用 macOS 原生内置的 `Apple Vision.framework`(8 线程并发),700 多帧图像在 20 秒内即可全量 OCR 识别完毕!*
### 3. 一键生成 CRM 资产大盘
```bash
# 运行通用结构化引擎
python scripts/structure_contacts.py data/contacts_raw.txt --csv data/contacts.csv --html data/contacts_crm.html
```
*(可选) 如果你有自己的行业场景词表,可传入 JSON 配置文件:*
```bash
python scripts/structure_contacts.py data/contacts_raw.txt --config config.example.json
```
---
## 🧩 命名模式抽取与打标建议 (Naming Schema & Best Practices)
本技能的抽取效果依赖于用户日常给微信好友打标和备注的习惯。推荐的命名规范为:
$$\text{原备注} = [\text{微信名/称谓}] + [\text{职位/头衔/机构/学校}] + [\text{场地/活动/商圈/圈子}] + [\text{时间/年份}]$$
### 提取维度说明:
- **微信名 / 称谓**:自动剥离地点、时间后的核心真实姓名或昵称。
- **机构 / 公司 / 学校**:识别主流高校与知名企业机构。
- **场地 / 圈子 / 活动**:识别商圈、酒店、展会、校友会及沙龙等场景。
- **城市 / 地区**:识别主要城市与省份。
- **时间**:支持 4 位年份(如 `2023`, `2024`)或 8 位具体日期(如 `20230501`)。
- **行业自动分类**:金融/投资、科技/AI/互联网、法律/法务/咨询、高校/科研、医疗健康、文化消费等。
> 📌 **建议**:
> 在日常添加好友时,养成在备注末尾加上「场景 + 年份」的习惯(例如:`张三 某某科技 某某峰会 2024`),就能随时一键生成详实的人脉资产图谱!
---
## 📊 产出物一览
1. 💻 **交互式 HTML CRM 看板** (`contacts_crm.html`):
- 极具现代感的 Dark Glassmorphism(暗黑玻璃拟态)设计。
- 内置多条件即时模糊搜索(输入姓名、场地、城市、行业即时响应)。
- 多维筛选下拉框(按行业、按城市一键过滤)。
2. 📈 **结构化 CSV 表格** (`contacts.csv`):
- UTF-8 BOM 编码,双击直接在 Excel / Apple Numbers 中打开,不乱码。
3. 💾 **标准 JSON 数据文件** (`contacts.json`):
- 便于导入 SQLite、DuckDB 或接入企业内部 Agenskill-card.md
## Description: Zero-risk, permission-free WeChat personal contacts extraction and structured CRM pipeline. Uses macOS native screen recording (Cmd+Shift+5), high-FPS FFmpeg frame extraction, Apple Silicon concurrent Vision OCR, and multi-dimensional rule-based NLP extraction (Name, Title, Org, Venue, City, Time) to generate Excel/CSV, JSON, and interactive HTML dashboards. This skill is ready for commercial/non-commercial use. ## Publisher: [emergencescience](https://clawhub.ai/user/emergencescience) ### License/Terms of Use: MIT ## Use Case: External users, developers, and operators use this skill to convert authorized WeChat contact-list screen recordings into structured CRM-ready contact records and dashboards. It supports local OCR extraction, deduplication, rule-based contact parsing, and export workflows for personal network analysis. ### Deployment Geography for Use: Global ## Known Risks and Mitigations: Risk: The skill bulk-extracts WeChat contacts into persistent raw text, CSV, JSON, and HTML CRM files. Mitigation: Use it only for contacts the user is authorized to process, keep generated outputs in restricted storage, and avoid shared or cloud-synced folders. Risk: The generated dashboard can expose contact data through remote scripts or fonts. Mitigation: Remove remote script and font dependencies before opening or sharing the dashboard, especially in sensitive environments. Risk: Contact values are rendered into the generated dashboard and may create injection exposure. Mitigation: Escape or sanitize contact values before rendering them in HTML and review generated dashboards before distribution. ## Reference(s): - [ClawHub Skill Page](https://clawhub.ai/emergencescience/skills/emergence-wechat-contact-crm) - [Server-resolved GitHub Repository](https://github.com/emergencescience/emergence-wechat-contact-crm) - [Server-resolved Source Commit](https://github.com/emergencescience/emergence-wechat-contact-crm/commit/bbcc6744be28008855be5c449e74ff158d343fbf) ## Skill Output: **Output Type(s):** [Text, Code, Shell commands, Configuration, Guidance, Files] **Output Format:** [Markdown guidance with shell commands; generated text, CSV, JSON, and HTML files] **Output Parameters:** [1D] **Other Properties Related to Output:** [Outputs can contain personal contact data and should be stored in restricted local locations.] ## Skill Version(s): 0.1.0 (source: ClawHub release metadata; artifact frontmatter reports 1.0.0) ## Ethical Considerations: Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
config.example.json
{
"cities": ["北京", "上海", "深圳", "广州", "杭州", "香港", "新加坡", "伦敦", "纽约"],
"roles": ["创始人", "合伙人", "CEO", "总监", "经理", "工程师", "研究员", "律师", "店长"],
"venues": ["某某酒店", "某某商圈", "某某峰会", "某某校友会", "某某俱乐部"],
"orgs": ["某某大学", "某某科技", "某某集团", "某某银行"]
}AionUi
Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!
activepieces
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
cherry-studio
AI productivity studio with smart chat, autonomous agents, and 300+ assistants.
CopilotKit
The Frontend for Agents & Generative UI. React + Angular
Machine-readable data
The same record, as JSON, for agents and crawlers.
{
"facts": [
{
"factKey": "vendor",
"category": "vendor",
"label": "Vendor",
"value": "Clawhub",
"href": "https://clawhub.ai/emergencescience/skills/emergence-wechat-contact-crm",
"sourceUrl": "https://clawhub.ai/emergencescience/skills/emergence-wechat-contact-crm",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-10T12:45:19.773Z",
"isPublic": true
},
{
"factKey": "protocols",
"category": "compatibility",
"label": "Protocol compatibility",
"value": "OpenClaw",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-emergencescience-emergence-wechat-contact-crm/contract",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-emergencescience-emergence-wechat-contact-crm/contract",
"sourceType": "contract",
"confidence": "medium",
"observedAt": "2026-10-10T12:45:19.773Z",
"isPublic": true
},
{
"factKey": "traction",
"category": "adoption",
"label": "Adoption signal",
"value": "1.4K downloads",
"href": "https://clawhub.ai/emergencescience/emergence-wechat-contact-crm",
"sourceUrl": "https://clawhub.ai/emergencescience/emergence-wechat-contact-crm",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-10T12:45:19.773Z",
"isPublic": true
},
{
"factKey": "latest_release",
"category": "release",
"label": "Latest release",
"value": "0.1.0",
"href": "https://clawhub.ai/emergencescience/emergence-wechat-contact-crm",
"sourceUrl": "https://clawhub.ai/emergencescience/emergence-wechat-contact-crm",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-08-17T12:03:34.881Z",
"isPublic": true
},
{
"factKey": "handshake_status",
"category": "security",
"label": "Handshake status",
"value": "UNKNOWN",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-emergencescience-emergence-wechat-contact-crm/trust",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-emergencescience-emergence-wechat-contact-crm/trust",
"sourceType": "trust",
"confidence": "medium",
"observedAt": null,
"isPublic": true
}
],
"events": [
{
"eventType": "release",
"title": "Release 0.1.0",
"description": "WeChat Contact List Video Extraction & CRM Analyzer 1.0.0 - Initial release of a macOS-based tool for extracting WeChat personal contacts via screen recording, ensuring zero-risk and no system permission requirements. - Utilizes FFmpeg for high-FPS frame extraction and Apple Vision OCR with concurrent processing for rapid text recognition. - Implements rule-based NLP to parse contact details (Name, Title, Organization, Venue, City, Time) into structured data. - Outputs results in Excel/CSV, JSON, and interactive HTML dashboards for CRM applications. - Includes detailed usage guide covering best practices for WeChat contact remarking, extraction performance metrics, and command-line steps for end-to-end operation.",
"href": "https://clawhub.ai/emergencescience/emergence-wechat-contact-crm",
"sourceUrl": "https://clawhub.ai/emergencescience/emergence-wechat-contact-crm",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-08-17T12:03:34.881Z",
"isPublic": true
}
]
}Record generated Oct 10, 2026.
