Lead Enrichment — Research Prospects in Seconds
Stop spending hours stalking LinkedIn. Let your agent do it.
Sales teams waste 6+ hours per week manually researching prospects. You Google their name, check LinkedIn, scroll their Twitter, hunt for their email, read their company's About page, search for recent news... and then do it all over again for the next lead.
Lead Enrichment automates all of it. Give your agent a name and company (or email, or LinkedIn URL), and get back a complete dossier: contact info, social profiles, bio, company intel, recent posts, news mentions, and AI-generated talking points.
The pain: Generic outreach gets ignored. Personalization takes forever. You're always behind quota.
The fix: Your agent researches 10 leads while you grab coffee. Rich profiles ready when you need them. Spend your time selling, not searching.
What You Get
For each lead, the enrichment pulls:
Personal Profile:
- - Full name, current title, company
- Professional bio/summary
- Profile photo URL
- Location
- Social media handles (LinkedIn, Twitter, GitHub, personal site)
Contact Discovery:
- - Likely email addresses (pattern-based + verification attempts)
- Public phone numbers (if available)
- Best channels for outreach
Company Context:
- - Company description, industry, size
- Funding stage, recent news
- Tech stack (for technical sales)
- Key decision makers
Intelligence & Timing:
- - Recent posts/articles (last 30 days)
- Job change signals
- Company news mentions
- Shared connections or interests
- Conference/event participation
AI-Generated Talking Points:
- - 3-5 personalized hooks based on their recent activity
- Common ground opportunities
- Relevant pain points to address
- Recommended opening lines
Setup
- 1. Run
scripts/setup.sh to initialize config - Edit
~/.config/lead-enrichment/config.json with preferences - No API keys required for basic enrichment (uses public sources)
- Optional: Add premium data sources (see config)
Scripts
| Script | Purpose |
|---|
| INLINECODE2 | Initialize config and data directories |
| INLINECODE3 |
Enrich a single lead (main script) |
|
scripts/batch.sh | Process multiple leads from CSV/JSON |
|
scripts/export.sh | Export enriched leads (JSON/MD/CSV) |
Usage
Single Lead
CODEBLOCK0
Batch Processing
CODEBLOCK1
Export Formats
CODEBLOCK2
Config
Config lives at ~/.config/lead-enrichment/config.json. See config.example.json for full schema.
Key sections:
enrichment.sources — Which data sources to check (all public by default):
- -
linkedin — Public profiles via search - INLINECODE9 — Social activity and bio
- INLINECODE10 — For technical leads
- INLINECODE11 — About pages, team directories
- INLINECODE12 — Recent mentions
- INLINECODE13 — Company funding (public data)
enrichment.depth — How thorough to be:
- -
quick — Basic profile only (name, title, LinkedIn, company) - INLINECODE15 — Above + social profiles + recent activity (default)
- INLINECODE16 — Above + news mentions + talking points + shared connections
output.format — Default output format (json/markdown/csv)
output.include — What to include in output:
- -
contact_info — Email attempts, phone - INLINECODE18 — All discovered links
- INLINECODE19 — Posts, articles (last 30 days)
- INLINECODE20 — Company description, size, funding
- INLINECODE21 — AI-generated personalization hooks
- INLINECODE22 — Source URLs for verification
talking_points.enabled — Generate AI talking points (requires Claude)
talking_points.style — Tone for suggestions (professional/friendly/bold)
privacy.respect_robots — Skip profiles with clear "no scraping" signals
privacy.store_locally — Cache enriched profiles (default: true)
Data Sources
All sources are public and free:
- 1. LinkedIn — Public profiles via search (no API, respects robots.txt)
- Twitter/X — Bio, recent tweets, follower count
- GitHub — For technical roles (repos, activity, README)
- Company websites — Team pages, About sections
- Google News — Recent mentions
- Crunchbase — Public company data (no API key needed for basic info)
- Common email patterns — firstname@company.com, f.lastname@company.com, etc.
Premium sources (optional, requires API keys):
- - Hunter.io — Email verification
- Clearbit — Enhanced company data
- Apollo — Direct contact info
Add API keys to ~/.clawdbot/secrets.env if you have them. Enrichment works fine without them.
Output Schema
Each enriched lead is saved as JSON:
CODEBLOCK3
Integration with Trawl
Lead Enrichment pairs perfectly with Trawl (autonomous lead gen):
CODEBLOCK4
Tips
Email Discovery:
- - Works best when you provide company domain
- Tries common patterns (first@company, f.last@company, etc.)
- Marks confidence level (high/medium/low)
- Does NOT spam or verify via email sends (respects privacy)
Talking Points:
- - Most valuable when enrichment depth = "deep"
- Requires recent activity data (posts, news)
- AI analyzes content for personalization hooks
- Style can be professional, friendly, or bold
Batch Processing:
- - Use
--parallel for speed (3-5 concurrent recommended) - Progress saved (resume if interrupted)
- Failed leads logged to INLINECODE25
Data Freshness:
- - Cached profiles expire after 30 days
- Force refresh with
--refresh flag - Social activity always fetched fresh
Use Cases
Sales Reps:
- - Research prospects before calls
- Personalize cold email sequences
- Find mutual connections or interests
Recruiters:
- - Assess candidate backgrounds
- Find contact info for passive candidates
- Check GitHub activity for technical roles
Partnerships:
- - Research potential partners
- Understand company context
- Find the right contact person
Investors:
- - Quick founder background checks
- Company traction signals
- Network mapping
Privacy & Ethics
This skill only uses publicly available data. It:
- - Respects robots.txt and rate limits
- Does NOT scrape private profiles or paywalled content
- Does NOT send verification emails (won't spam your leads)
- Does NOT store data if privacy.store_locally = false
- Provides source URLs for transparency
Be a human: Just because you CAN enrich someone doesn't mean you should spam them. Use this for genuine, personalized outreach.
Data Storage
Enriched leads are stored at ~/.config/lead-enrichment/data/leads/:
CODEBLOCK5
FAQ
Q: Is this legal?
A: Yes. All data is publicly available. We respect robots.txt and rate limits.
Q: How accurate are the emails?
A: Pattern-based = 60-80% accuracy. Verified (if you add Hunter.io key) = 95%+.
Q: Can I enrich 1000 leads?
A: Yes via batch.sh. Expect ~30 sec per lead (deep mode). That's 8 hours for 1000. Run overnight.
Q: Does this work for non-US leads?
A: Yes. LinkedIn and Twitter are global. Some data sources are US-biased.
Q: Will this get me blocked by LinkedIn?
A: No. We use search (public), not scraping. Rate-limited and respectful.
What's Next
Ideas for future versions:
- - Chrome extension (enrich while browsing LinkedIn)
- CRM integrations (auto-enrich on lead create)
- Slack bot (enrich on-demand from Slack)
- Email warmup integration (find + verify + warm sequence)
- Mutual connection finder (via agent networks)
- Real-time alerts (when a lead changes jobs)
Stop researching. Start selling.
Feed your agent a list of names. Get back a stack of dossiers. Personalize every message. Close more deals.
That's Lead Enrichment.
潜在客户信息丰富 — 数秒内完成客户研究
别再花几小时刷领英了。让你的智能代理去做吧。
销售团队每周花费超过6小时手动研究潜在客户。你在谷歌搜索他们的名字,查看领英,翻阅推特,寻找邮箱,阅读公司关于页面,搜索最新新闻……然后对下一个潜在客户重复整个过程。
潜在客户信息丰富功能将这一切自动化。只需向你的智能代理提供姓名和公司(或邮箱、领英链接),即可获得完整档案:联系方式、社交资料、个人简介、公司情报、近期动态、新闻报道以及AI生成的谈话要点。
痛点: 千篇一律的推广被忽视。个性化定制耗时漫长。你总是落后于业绩指标。
解决方案: 你喝杯咖啡的功夫,智能代理就能研究10个潜在客户。需要时,丰富的档案随时就绪。把时间花在销售上,而不是搜索上。
你将获得
针对每个潜在客户,信息丰富功能会提取:
个人资料:
- - 全名、现任职位、公司
- 职业简介/摘要
- 头像链接
- 所在地
- 社交媒体账号(领英、推特、GitHub、个人网站)
联系方式发现:
- - 可能的邮箱地址(基于模式 + 验证尝试)
- 公开电话号码(如有)
- 最佳联系渠道
公司背景:
- - 公司描述、行业、规模
- 融资阶段、近期新闻
- 技术栈(用于技术销售)
- 关键决策者
情报与时机:
- - 近期动态/文章(最近30天)
- 工作变动信号
- 公司新闻提及
- 共同联系或兴趣
- 会议/活动参与
AI生成的谈话要点:
- - 基于其近期活动的3-5个个性化切入点
- 共同点机会
- 需解决的相关痛点
- 推荐的开场白
设置
- 1. 运行 scripts/setup.sh 初始化配置
- 编辑 ~/.config/lead-enrichment/config.json 设置偏好
- 基础信息丰富无需API密钥(使用公开来源)
- 可选:添加高级数据源(见配置)
脚本
| 脚本 | 用途 |
|---|
| scripts/setup.sh | 初始化配置和数据目录 |
| scripts/enrich.sh |
丰富单个潜在客户信息(主脚本) |
| scripts/batch.sh | 批量处理CSV/JSON中的多个潜在客户 |
| scripts/export.sh | 导出丰富后的潜在客户信息(JSON/MD/CSV) |
使用方法
单个潜在客户
bash
按姓名 + 公司
./scripts/enrich.sh --name Sarah Chen --company Acme Corp
按邮箱
./scripts/enrich.sh --email sarah@acmecorp.com
按领英链接
./scripts/enrich.sh --linkedin https://linkedin.com/in/sarahchen
输出到文件
./scripts/enrich.sh --name Sarah Chen --company Acme Corp --output sarah-chen.json
附带谈话要点
./scripts/enrich.sh --name Sarah Chen --company Acme Corp --talking-points
批量处理
bash
从CSV(列:name, company, email, linkedin_url)
./scripts/batch.sh --input leads.csv --output enriched/
从JSON数组
./scripts/batch.sh --input leads.json --output enriched/
并发处理
./scripts/batch.sh --input leads.csv --parallel 3
导出格式
bash
导出为JSON(默认)
./scripts/export.sh --format json enriched/*.json > leads.json
导出为Markdown(可读)
./scripts/export.sh --format markdown enriched/*.json > leads.md
导出为CSV(CRM导入)
./scripts/export.sh --format csv enriched/*.json > leads.csv
管道输出到CRM
./scripts/export.sh --format json enriched/*.json | \
curl -X POST https://your-crm.com/api/leads -d @-
配置
配置文件位于 ~/.config/lead-enrichment/config.json。完整模式参见 config.example.json。
关键部分:
enrichment.sources — 要检查的数据源(默认全部公开):
- - linkedin — 通过搜索获取公开资料
- twitter — 社交活动和简介
- github — 针对技术类潜在客户
- company_website — 关于页面、团队目录
- news — 近期提及
- crunchbase — 公司融资(公开数据)
enrichment.depth — 详细程度:
- - quick — 仅基础资料(姓名、职位、领英、公司)
- standard — 上述内容 + 社交资料 + 近期活动(默认)
- deep — 上述内容 + 新闻提及 + 谈话要点 + 共同联系
output.format — 默认输出格式(json/markdown/csv)
output.include — 输出包含内容:
- - contactinfo — 邮箱尝试、电话
- socialprofiles — 所有发现的链接
- recentactivity — 动态、文章(最近30天)
- companyintel — 公司描述、规模、融资
- talkingpoints — AI生成的个性化切入点
- rawsources — 用于验证的源链接
talking_points.enabled — 生成AI谈话要点(需Claude)
talking_points.style — 建议语气(专业/友好/大胆)
privacy.respect_robots — 跳过明确标记禁止抓取的资料
privacy.store_locally — 缓存丰富后的资料(默认:true)
数据源
所有数据源均为公开且免费:
- 1. 领英 — 通过搜索获取公开资料(无API,遵守robots.txt)
- 推特/X — 简介、近期推文、粉丝数
- GitHub — 针对技术岗位(仓库、活动、README)
- 公司网站 — 团队页面、关于部分
- 谷歌新闻 — 近期提及
- Crunchbase — 公开公司数据(基础信息无需API密钥)
- 常见邮箱模式 — firstname@company.com、f.lastname@company.com等
高级数据源(可选,需API密钥):
- - Hunter.io — 邮箱验证
- Clearbit — 增强公司数据
- Apollo — 直接联系方式
如有API密钥,请添加到 ~/.clawdbot/secrets.env。没有密钥,信息丰富功能也能正常运行。
输出模式
每个丰富后的潜在客户保存为JSON:
json
{
lead_id: sarah-chen-acme-corp,
enriched_at: 2025-01-29T10:30:00Z,
input: {
name: Sarah Chen,
company: Acme Corp
},
profile: {
full_name: Sarah Chen,
title: VP of Engineering,
company: Acme Corp,
location: San Francisco, CA,
bio: Building the future of...,
photo_url: https://...,
social_profiles: {
linkedin: https://linkedin.com/in/sarahchen,
twitter: https://twitter.com/sarahchen,
github: https://github.com/sarahchen,
personal_site: https://sarahchen.com
}
},
contact: {
emails: [
{ address: sarah@acmecorp.com, confidence: 0.85, verified: false },
{ address: s.chen@acmecorp.com, confidence: 0.60, verified: false }
],
phones: [],
preferred_channel: email
},
company: {
name: Acme Corp,
domain: acmecorp.com,
industry: SaaS,
size: 51-200 employees,
description: AI-powered...,
funding: Series B ($25M),
tech_stack: [React, Node.js, AWS],
recent_news: [
{
title: Acme Corp raises $25M...,
url: https://...,
date: 2025-01-15
}
]
},
intelligence: {
recent_activity: [
{
type: twitter_post,
content: Excited to announce...,
url: https://...,
date: 2025-01-20
}
],
jobchangesignal: false,