Skip to main content
Glama
72bpm-leadgen-mcp-server

72bpm-leadgen-mcp-server

72BPM Lead-Gen MCP Server

An MCP server that turns discovery → enrichment → scoring → outreach into callable tools, scoped to 72BPM's four practice areas:

  • SaaS Product Development

  • E-mobility (EV) Applications & Infrastructure

  • IoT Platforms

  • AI Agent Creation

Why no LinkedIn API/scraper?

LinkedIn's Terms of Service prohibit automated scraping, and LinkedIn actively detects and blocks it — both technically (rate limits, IP bans) and legally. Rather than build something that breaks on day one and creates legal exposure, this server:

  • Crawls companies' own public websites directly (careers pages, blog posts, about pages) — fully fair game, standard practice.

  • Uses a real search engine (Brave Search API) to read back publicly indexed snippets — including of LinkedIn company/people pages — the same way a human researcher googling site:linkedin.com/company acme would. This is a search-engine query, not a scrape of linkedin.com itself.

  • Reads public job postings the same way.

If your team later gets access to LinkedIn's official Talent/Sales Navigator API (via their Partner Program) or a compliant data provider, swap src/services/searchProvider.ts — every tool depends only on that one file's webSearch() function.

Related MCP server: Commonality MCP Server

Tools

Tool

Purpose

leadgen_discover_companies

Search-based candidate discovery for one category + region

leadgen_enrich_company

Crawl a company's site + indexed LinkedIn/job signal, matched against all 4 categories

leadgen_score_lead

Weighted 0-100 fit score + tier (hot/warm/cold) + confidence for one category

leadgen_find_contact

Search-based candidate engineering contact (CTO/VP Eng/etc.) — unverified, always double-check before outreach

leadgen_generate_pitch

Drafts 2 outreach email variants using the value-lead framework

leadgen_save_lead / leadgen_list_leads / leadgen_update_lead

Simple pipeline tracking (JSON file by default)

Typical workflow: discover → enrich → score → find_contact → generate_pitch → save_lead, then update_lead as deals progress.

Setup

npm install
cp .env.example .env
# edit .env and set BRAVE_API_KEY (free tier: https://brave.com/search/api/)
npm run build

Run locally (stdio) — for Claude Desktop / Claude Code

npm start

Add to your Claude Desktop / Claude Code MCP config:

{
  "mcpServers": {
    "72bpm-leadgen": {
      "command": "node",
      "args": ["/absolute/path/to/72bpm-leadgen-mcp-server/dist/index.js"],
      "env": {
        "BRAVE_API_KEY": "your-key-here"
      }
    }
  }
}

Run as a hosted server (streamable HTTP) — for team-wide access

TRANSPORT=http PORT=3000 npm start
# or: npm run start:http

Exposes POST /mcp (the MCP endpoint) and GET /health. Point your team's MCP clients at https://your-host/mcp. Deploy behind your normal reverse proxy/TLS termination — this server does not handle auth itself, so put it behind your existing auth layer if exposing it beyond your VPN.

Customizing the ICP signals

src/constants.ts holds the weighted signal dictionary per category (ICP_SIGNALS). These are the phrases the scoring engine looks for and how many points each is worth. Tune them as you learn what actually converts — e.g. if "battery swapping" mentions never turn into real deals, drop its weight; if a new signal phrase keeps showing up in your best deals, add it.

Tier thresholds (tierFromScore) and confidence thresholds (confidenceFromSourceCount) live in the same file.

Customizing outreach voice

src/tools/pitch.ts has one AUTHORITY_LINES entry per category — the "parallel domain authority" sentence 72BPM's pitches lean on. Edit these to match your team's actual case studies as they accumulate; specific, verifiable claims outperform generic ones.

Known limitations (by design, not oversights)

  • No paid enrichment APIs (Apollo/Clearbit/Crunchbase/BuiltWith) — you said you don't have those yet. The signal quality here is real but bounded by what's publicly crawlable. If you add API keys later, this is the natural next upgrade — add a new services/ client and feed its text into matchSignals() alongside the site crawl.

  • Contact finding is unverified — leadgen_find_contact parses search snippet text, which is sometimes ambiguous. It's a research shortcut, not a verified contact database. Always confirm before sending outreach.

  • Pipeline storage is a JSON file — fine for one person or a small team running the stdio server locally. For the hosted HTTP deployment with multiple concurrent users, swap src/services/storage.ts for a real database (Postgres/SQLite) — every other tool only depends on the four functions that file exports.

  • Brave Search rate limits apply — the free tier is fine for exploration but will need a paid tier for high-volume prospecting.

Project structure

src/
├── index.ts              # Entry point, dual transport (stdio/HTTP)
├── constants.ts           # ICP signal weights, category labels, thresholds
├── types.ts                # Shared TypeScript types
├── services/
│   ├── searchProvider.ts   # Brave Search API wrapper (swap for another provider here)
│   ├── webCrawler.ts        # Company-site crawler (robots.txt-aware)
│   ├── companyIntel.ts       # Combines crawl + search into cached text bundles
│   ├── scoring.ts             # Weighted signal matching + scoring
│   └── storage.ts              # JSON-file lead pipeline (swap for a DB for hosted use)
└── tools/
    ├── discover.ts
    ├── enrich.ts
    ├── score.ts
    ├── contact.ts
    ├── pitch.ts
    └── pipeline.ts

Available Tools

8 tools
leadgen_discover_companiesDiscover Candidate CompaniesA
Read-only

Searches the public web for companies showing technical ICP signals for one of 72BPM's four practice areas, and returns a de-duplicated, signal-ranked candidate list.

This tool does NOT scrape LinkedIn or any social platform directly. It queries a real search engine (Brave Search) the same way a human researcher would, and reads back only the public snippet text already indexed. Use leadgen_enrich_company next to go deeper on any promising candidate.

Args:

  • category ('saas' | 'e_mobility' | 'iot' | 'ai_agents'): which practice area to prospect for

  • region (string, optional): geographic focus, e.g. "India", "UAE"

  • extra_keywords (string[], optional): up to 5 extra terms to narrow the search

  • max_results (number): 1-20, default 8

Returns: JSON with schema: { "category": string, "category_label": string, "queries_used": string[], "candidates": [ { "name": string, // best-guess company/page name from search result title "domain": string | null, // extracted domain, use with leadgen_enrich_company "matched_signals": string[],// ICP phrases found in the search snippet itself "sources": [{ "title": string, "url": string, "snippet": string }] } ], "count": number }

Examples:

  • Use when: "Find e-mobility companies in India working on payment sync" -> category="e_mobility", region="India", extra_keywords=["payment sync"]

  • Use when: "Who's building multi-agent systems in the UAE" -> category="ai_agents", region="UAE"

  • Don't use when: you already have a company name and just want its tech profile (use leadgen_enrich_company instead)

Error Handling:

  • Returns an error message if BRAVE_API_KEY is not configured, with a link to get one

  • Returns "No candidates found" with the queries tried if the search turned up nothing usable

Inputs: category (required), region, max_results, extra_keywords.

leadgen_enrich_companyEnrich Company Technical ProfileA
Read-only

Builds a technical profile for a specific company by crawling its own public website (homepage, /careers, /about, /blog, /engineering) and cross-referencing an indexed LinkedIn snippet plus public job postings — all via legitimate, ToS-compliant sources (no LinkedIn scraping).

Returns which ICP signal phrases were found, for ALL FOUR practice areas at once, so you can see if a company is a multi-category fit (e.g. an IoT hardware company also hiring for LLM-based ops automation).

Args:

  • company_name (string): required

  • domain (string, optional): website domain; enrichment is much richer with it, but LinkedIn/job-posting signal still works without it

Returns: JSON with schema: { "company_name": string, "domain": string | null, "crawled_pages": string[], // URLs successfully crawled "crawl_errors": string[], // pages skipped/failed, with reason "linkedin_snippet": { "title": string, "url": string, "snippet": string } | null, "job_posting_hits": [{ "title": string, "url": string, "snippet": string }], "signals_by_category": { "": { "matched": [{ "phrase": string, "weight": number, "found_in": string[] }], "missing": string[] } } }

Examples:

  • Use when: "What's ElectreeFi's tech stack look like?" -> company_name="ElectreeFi", domain="electreefi.com"

  • Use when: you have a candidate from leadgen_discover_companies and want the full picture before scoring

  • Don't use when: you just want a single category's numeric fit score (use leadgen_score_lead — it's cheaper and more direct)

Error Handling:

  • If domain is unreachable, crawl_errors will explain why but the tool still returns LinkedIn/job-posting signal

  • Returns an error if BRAVE_API_KEY is missing and no domain was provided (nothing to enrich from)

Inputs: company_name (required), domain.

leadgen_find_contactFind Candidate Engineering ContactA
Read-only

Searches public, search-engine-indexed snippets (not a LinkedIn scrape) for a likely engineering decision-maker at a company — CTO, VP Engineering, Head of Engineering, etc.

IMPORTANT: results are UNVERIFIED candidates parsed from search snippet text. They are a starting point for manual confirmation (open the profile URL yourself, or verify via a mutual connection / company site team page), not a guaranteed accurate contact. Never send outreach based solely on this tool's output without a human sanity-check.

Args:

  • company_name (string): required

  • role_hint (string, optional): a specific title to prioritize, e.g. "Head of Engineering". Defaults to trying CTO, VP Engineering, Head of Engineering, Engineering Lead, Director of Engineering.

Returns: JSON with schema: { "company_name": string, "candidates": [ { "name_guess": string | null, "title_guess": string | null, "profile_url": string | null, "snippet": string, "verified": false // always false — see IMPORTANT note above } ], "count": number }

Examples:

  • Use when: "Who leads engineering at ElectreeFi?" -> company_name="ElectreeFi"

  • Use when: "Find their Head of Data" -> company_name="Acme Corp", role_hint="Head of Data"

  • Don't use when: you need a verified email address (this tool does not find or guess emails)

Error Handling:

  • Returns "No candidates found" if search turns up nothing matching common engineering-leadership titles

  • Returns an error if BRAVE_API_KEY is not configured

Inputs: company_name (required), role_hint.

leadgen_generate_pitchGenerate Outreach PitchA
Read-onlyIdempotent

Drafts two personalized outreach email variants for a qualified lead, following 72BPM's value-lead framework: lead with a specific technical friction point (not a sales pitch), show parallel domain authority for the relevant practice area, and close with a low-friction peer-to-peer call to action.

This tool does NOT send anything — it only drafts text for you to review and send yourself.

Args:

  • company_name (string): required

  • category ('saas' | 'e_mobility' | 'iot' | 'ai_agents'): required, determines which authority/experience line is used

  • friction_point (string): required — the specific technical gap to lead with. Pull this from leadgen_enrich_company's matched_signals or crawled content; generic friction points make for generic (worse) pitches

  • contact_name (string, optional): personalizes the greeting

  • contact_title (string, optional): for your own reference / future use

  • sender_name (string, default "[Your name]"): signs the email

Returns: JSON with schema: { "company_name": string, "category": string, "variants": [ { "label": string, "subject": string, "body": string } ] }

Examples:

  • Use when: you've scored a lead as "hot" and identified its top missing/matched signal, and want ready-to-edit email drafts

  • Don't use when: you haven't identified a real, specific friction point yet — generic pitches convert poorly; go back to leadgen_enrich_company first

Error Handling:

  • Returns a validation error if friction_point is missing or too short — this tool refuses to draft a pitch with nothing specific to say

Inputs: category (required), company_name (required), friction_point (required), sender_name, +2 more.

leadgen_list_leadsList Pipeline LeadsA
Read-onlyIdempotent

Lists saved leads from the pipeline, optionally filtered by category, tier, or status.

Args:

  • category, tier, status (all optional): filter the results

Returns: JSON: { "leads": StoredLead[], "count": number } where StoredLead has fields: id, companyName, domain, category, region, score, tier, status, notes, contact, createdAt, updatedAt

Examples:

  • Use when: "Show me all hot e-mobility leads" -> category="e_mobility", tier="hot"

  • Use when: "What's still uncontacted?" -> status="new"

Inputs: tier, status, category.

leadgen_save_leadSave Lead to PipelineA

Persists a qualified lead to the local pipeline store (a JSON file under ./data by default; see README for swapping in a real database for hosted/multi-user deployments).

Args:

  • company_name (string): required

  • domain, region (string, optional)

  • category ('saas' | 'e_mobility' | 'iot' | 'ai_agents'): required

  • score (number 0-100, optional): from leadgen_score_lead

  • tier ('hot' | 'warm' | 'cold', optional): from leadgen_score_lead

  • notes (string, optional)

  • contact_name, contact_title, contact_profile_url (string, optional): from leadgen_find_contact

Returns: the saved lead record including its generated "id", which you'll need for leadgen_update_lead.

Examples:

  • Use when: you've scored a company as hot/warm and want to track it for follow-up

  • Don't use when: you're just browsing candidates (save only once you're tracking it for real)

Inputs: category (required), company_name (required), tier, notes, +6 more.

leadgen_score_leadScore Lead FitA
Read-only

Computes a weighted 0-100 ICP fit score for one company against one 72BPM practice area, with a tier (hot/warm/cold) and a confidence level based on how many independent public sources corroborated the signal.

Internally this reuses the same public-source gathering as leadgen_enrich_company (and shares its short-lived cache, so calling enrich then score on the same company won't double the network calls).

Args:

  • company_name (string): required

  • domain (string, optional): website domain — scoring is more reliable with it

  • category ('saas' | 'e_mobility' | 'iot' | 'ai_agents'): required, which practice area to score against

Returns: JSON with schema: { "company_name": string, "category": string, "category_label": string, "score": number, // 0-100 "tier": "hot" | "warm" | "cold", "confidence": "low" | "medium" | "high", // based on # of independent sources with matches "matched_signals": [{ "phrase": string, "weight": number, "found_in": string[] }], "missing_high_value_signals": string[], // top signals NOT found — good discovery-call questions "rationale": string }

Tier thresholds: score >= 55 = hot, >= 30 = warm, else cold. Confidence: 3+ independent sources = high, 2 = medium, 0-1 = low. Treat "low confidence + hot tier" as promising but unverified — worth a human look before outreach.

Examples:

  • Use when: "Is Acme Corp a good IoT lead?" -> company_name="Acme Corp", domain="acme.com", category="iot"

  • Don't use when: you want the full multi-category breakdown (use leadgen_enrich_company instead)

Error Handling:

  • Returns an error if BRAVE_API_KEY is missing and domain gathers nothing

  • Returns score=0, tier="cold", confidence="low" (not an error) if the company has no public signal at all — that's a valid, informative result

Inputs: category (required), company_name (required), domain.

leadgen_update_leadUpdate Pipeline LeadA
Idempotent

Updates a saved lead's status and/or notes — e.g. after sending outreach or getting a reply.

Args:

  • lead_id (string, uuid): required, from leadgen_save_lead or leadgen_list_leads

  • status ('new'|'researching'|'contacted'|'replied'|'qualified'|'disqualified', optional)

  • notes (string, optional)

At least one of status or notes must be provided.

Returns: the updated lead record, or an error if lead_id doesn't exist.

Inputs: lead_id (required), notes, status.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.6/5.0

    Scored across 8 tools

    Disambiguation5/5

    Each tool has a distinct phase in the lead-gen workflow: discover finds candidates, enrich profiles a company, score quantifies fit, find_contact locates people, generate_pitch writes emails, and save/list/update manage the pipeline. Even the similar enrich vs. score are clearly differentiated by multi-category vs. single-category output.

    Naming Consistency5/5

    All tools follow the exact pattern leadgen_<verb>_<noun> with clear, consistent verbs (discover, enrich, score, find, generate, save, list, update). The prefix unifies the namespace and the verb_noun structure makes each tool's purpose predictable.

    Tool Count5/5

    8 tools is well-scoped for a specialized lead-generation workflow. Each tool covers a necessary step without redundancy or bloat, and the count is within the ideal range for a focused MCP server.

    Completeness4/5

    The lifecycle is nearly complete: discovery, enrichment, scoring, contact finding, pitch generation, and pipeline CRUD (save/list/update). The only notable gap is the absence of a delete/archive tool for leads, though the status field can handle disqualification, so agents can work around this.

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      A
      quality
      D
      maintenance
      AI-powered B2B outbound sales automation pipeline exposed as an MCP server that transforms natural language goals into qualified sales intelligence and personalized emails.
      9
      -
    • F
      license
      Not graded
      quality
      D
      maintenance
      Custom MCP server for HubSpot CRM providing ICP segmentation, scoring, duplicate detection, Data Quality Score, and lifecycle stage automation. Exposes 39 tools to enhance HubSpot's native capabilities.
      -