Scenario Guide

AI Lead Generation & Sales Prospecting with Agent Skills

Sales prospecting has historically been the most time-intensive part of the revenue cycle — hours spent on LinkedIn, manually enriching spreadsheets, and writing emails that feel personal but are templated. AI agent skills compress this into a fully automated pipeline: the agent identifies prospects matching your ICP, enriches them with firmographic and technographic data, scores them, drafts personalised outreach, and logs everything to your CRM — while you focus on closing. This guide covers the five essential lead generation skills and how to chain them into a production-ready prospecting workflow.

Table of Contents

  1. 1. What Is AI Lead Generation
  2. 2. Top 5 Lead Gen Agent Skills
  3. 3. Step-by-Step Setup
  4. 4. Prospecting Workflow Example
  5. 5. Comparison Table
  6. 6. FAQ (7 questions)
  7. 7. Related Resources

What Is AI Lead Generation with Agent Skills

AI lead generation with agent skills is the practice of automating the entire top-of-funnel sales process using an AI assistant that orchestrates specialised MCP tools. Unlike traditional CRM automation — which only moves data between predefined fields — agent-driven prospecting can reason about prospect fit, synthesise public signals into personalised messaging, and make decisions about who to contact and when.

The workflow begins with prospect identification: the agent searches for companies and individuals matching your ideal customer profile using Brave Search and LinkedIn data. It then enriches each record with firmographic, technographic, and intent data before applying a scoring model to prioritise who to contact first. The highest-scoring prospects receive personalised outreach emails drafted by the agent based on their specific context — a recent funding round, a relevant job posting, or a product announcement — rather than generic templates.

Because the agent logs every action to the CRM Skill, the pipeline is fully auditable. Sales managers can see exactly which prospects were identified, how they were scored, what outreach was sent, and what the response rate is — providing the data needed to continuously refine the ICP and scoring model.

Top 5 Lead Generation Agent Skills

These five skills address every stage of the prospecting pipeline, from initial identification through to CRM-tracked follow-up.

LinkedIn Scraper Skill

Medium

Community

Extracts structured profile and company data from LinkedIn using a headless browser session. Retrieves job titles, company size, industry, seniority, mutual connections, and recent activity signals that indicate purchase intent or hiring momentum.

Best for: B2B prospecting, ICP matching, account-based marketing lists

mcp-linkedin-scraper

Setup time: 10 min

Brave Search MCP

Low

Brave

Privacy-first search API that surfaces company news, product launches, funding announcements, and job postings — all strong intent signals. Use it to trigger outreach at the moment a prospect is most likely to engage with a relevant offer.

Best for: Trigger-based prospecting, news monitoring, company research enrichment

@modelcontextprotocol/server-brave-search

Setup time: 2 min

CRM Skill

Medium

Community (HubSpot / Salesforce)

Bidirectional integration with HubSpot or Salesforce CRM. Creates contacts, companies, and deals; updates lead scores; logs activities; and reads existing pipeline state so the agent can avoid re-contacting prospects already in an active sequence.

Best for: Lead creation, pipeline management, duplicate prevention, activity logging

mcp-hubspot-crm

Setup time: 10 min

Email Skill

Low

Community

Composes and sends personalised outreach emails via SendGrid or Resend. Supports dynamic personalisation tokens, multi-step sequences, open and click tracking, and automatic follow-up scheduling based on prospect engagement behaviour.

Best for: Cold outreach, follow-up sequences, transactional lead nurture

mcp-email-outreach

Setup time: 5 min

Enrichment Skill

Low

Community (Clearbit / Apollo)

Enriches a prospect's email address or domain with firmographic and technographic data: company revenue, headcount, tech stack, funding round, industry codes, and key decision-maker contacts — dramatically improving lead scoring accuracy.

Best for: Lead scoring, ICP filtering, personalisation data for outreach copy

mcp-lead-enrichment

Setup time: 5 min

Step-by-Step Setup

Configure all five skills in your MCP settings file and connect your CRM and email provider credentials.

Step 1: Add Skills to MCP Config

{
  "mcpServers": {
    "brave-search": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-brave-search"],
      "env": { "BRAVE_API_KEY": "$BRAVE_API_KEY" }
    },
    "linkedin-scraper": {
      "command": "npx",
      "args": ["-y", "mcp-linkedin-scraper"]
    },
    "crm": {
      "command": "npx",
      "args": ["-y", "mcp-hubspot-crm"],
      "env": { "HUBSPOT_API_KEY": "$HUBSPOT_API_KEY" }
    },
    "email": {
      "command": "npx",
      "args": ["-y", "mcp-email-outreach"],
      "env": { "SENDGRID_API_KEY": "$SENDGRID_API_KEY" }
    },
    "enrichment": {
      "command": "npx",
      "args": ["-y", "mcp-lead-enrichment"],
      "env": { "CLEARBIT_API_KEY": "$CLEARBIT_API_KEY" }
    }
  }
}

Step 2: Define Your ICP

Describe your ideal customer profile in plain language. The agent uses this as the scoring baseline: "Our ICP is B2B SaaS companies with 50–500 employees, Series A or B funded, using AWS or Vercel, and actively hiring for engineering roles — which indicates they have both the budget and technical need for our product."

Step 3: Run a Test Prospecting Session

Start with a small batch to validate the workflow: "Find 10 companies that match our ICP, enrich their data, score them, and create contacts in HubSpot for the top 5. Do not send any emails yet."

Prospecting Workflow: Identify to Track

  1. Identify prospects — Brave Search finds companies matching ICP signals; LinkedIn Scraper retrieves contact details.
  2. Enrich data — Enrichment Skill adds firmographic, technographic, and funding data to each prospect.
  3. Score leads — Agent applies ICP scoring model and ranks prospects by fit and intent signal strength.
  4. Outreach — Email Skill sends personalised sequences to top-scored prospects with company-specific opening lines.
  5. Track — CRM Skill logs all activity, updates lead scores on reply, and schedules follow-ups based on engagement.

Comparison Table

Skill responsibilities and data sources across the lead generation pipeline.

SkillPipeline StageData SourcePaid APISetup
LinkedIn ScraperIdentifyLinkedIn profilesNo (ToS risk)10 min
Brave Search MCPResearch / TriggerWeb search indexYes (2k/mo free)2 min
CRM SkillLog / TrackHubSpot / SalesforceYes10 min
Email SkillOutreachSendGrid / ResendYes (free tier)5 min
Enrichment SkillEnrich / ScoreClearbit / ApolloYes5 min

Frequently Asked Questions

What is AI lead generation with agent skills?

AI lead generation with agent skills means automating the full sales prospecting cycle — identifying potential buyers, enriching their data, scoring them against your ideal customer profile, personalising outreach, and tracking responses — through an AI agent that orchestrates specialised MCP skills. The agent replaces the manual process of searching LinkedIn, copy-pasting data into spreadsheets, and writing individual emails, compressing a multi-hour prospecting session into a minutes-long automated workflow.

Is LinkedIn scraping legal and within LinkedIn's terms of service?

LinkedIn's User Agreement prohibits automated scraping of its platform. This skill is documented for educational and research purposes. For production use cases, the recommended approach is to use LinkedIn's official Sales Navigator API or a compliant data provider like Apollo.io, Lusha, or Clearbit that has licensed the underlying data. These providers offer the same structured prospect data through legitimate API access covered by the Enrichment Skill.

How does the agent score and prioritise leads?

Lead scoring in an agent-driven workflow combines firmographic signals from the Enrichment Skill (company size, industry, tech stack, funding stage) with behavioural signals from Brave Search (recent hiring, product launches, news mentions) and CRM history (previous interactions, deal stage, last contact date). The agent applies a weighted scoring model — which you define in natural language or as a configuration file — and ranks prospects by score before initiating outreach sequences.

Can the Email Skill personalise outreach at scale without sounding generic?

Yes, when the Enrichment Skill and Brave Search MCP provide rich context. The agent draws on company-specific signals — a recent funding round, a job posting for a role your product addresses, a news mention — to craft an opening line that references something genuinely relevant to the prospect. This approach consistently outperforms generic templates because the personalisation is grounded in real data rather than inserted tokens like {{FirstName}}. The agent can generate and send 50–100 personalised emails per hour while maintaining distinct opening lines for each.

How does the CRM Skill prevent duplicate outreach?

Before creating a new contact or enrolling a prospect in a sequence, the CRM Skill queries the existing database for matching email addresses and company domains. If a record exists with an active deal, an open sequence, or a "Do Not Contact" flag, the agent skips the prospect and logs the reason. This deduplication step runs automatically as part of the workflow and prevents the embarrassing situation of cold-emailing a prospect who is already in active contract negotiations.

What sending limits should I set to avoid email spam flags?

Cold outreach deliverability degrades quickly above certain volume thresholds. Recommended limits: maximum 50 cold emails per day per sending domain, minimum 5-minute intervals between sends, and a 3-day minimum between follow-ups to the same prospect. The Email Skill supports configurable rate limiting and sending windows (e.g., send only between 9 AM and 5 PM in the prospect's timezone). Always warm up new sending domains for at least 2 weeks before scaling to full volume.

How do I measure the ROI of an AI lead generation workflow?

The primary metrics to track are time-to-pipeline (how long from identifying a prospect to creating an opportunity in CRM), sequence reply rate (industry benchmark: 5–15% for well-targeted cold outreach), and pipeline generated per hour of agent runtime. Compare these against your previous manual prospecting baseline. Most teams see a 4–8x improvement in leads researched per hour and a 2–3x improvement in personalisation quality when agent enrichment replaces manual research.