Loading repository data…
vinkius-labs / repository
AI agent crew that generates qualified B2B lead lists with enriched contacts and personalized outreach using CrewAI + 5 MCP servers + Google Gemini.
A transparent discovery signal based on current public GitHub metadata.
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
A multi-agent system that generates qualified B2B lead lists with enriched contact data and personalized outreach templates — powered by CrewAI, Google Gemini, and 5 production MCP servers from the Vinkius AI Gateway.
Sales teams spend an average of 21% of their day researching prospects. An SDR manually checking LinkedIn, searching Crunchbase for funding data, running Hunter.io for email addresses, and then writing a personalized email for each lead — that is 2+ hours per 10 leads, and the quality degrades with fatigue.
This project automates the entire pipeline. Three AI agents handle company research, contact enrichment, and outreach personalization in a single run. You define your Ideal Customer Profile, and the system delivers a qualified lead list with verified emails and ready-to-send outreach templates.
The key differentiator is that the agents access live data through the Model Context Protocol (MCP) — they query Apollo.io for firmographics, Crunchbase for funding data, Hunter for verified emails, LinkedIn for professional profiles, and Exa AI for recent company signals. No stale data. No guessing.
The first agent acts as a senior sales researcher. It uses two MCP servers:
Given your ICP (industry, company size, product fit), it finds 10 qualified companies, scores each for fit from 1-10, and provides reasoning for every score.
The second agent finds decision-makers at each qualified company. It uses two MCP servers:
It targets the specific role you define (VP of Engineering, Head of Product, CFO) and prioritizes email accuracy over speed.
Selected from shared topics, language and repository description—not editorial ratings.
arjunprabhulal /
An AI-powered travel planning application that uses Gemini LLM and CrewAI to automate trip planning with multiple specialized AI agents. This project demonstrates agentic AI collaboration to search flights, find hotels, and generate personalized travel itineraries.
57/100 healthadam-eques /
Multi-agent AI crew that researches a stock and writes a cited investment brief.
50/100 healthThe third agent creates outreach templates. It uses one MCP server:
Each email references something specific about the prospect's company — a recent funding round, a hiring trend, a technology adoption — connecting it to your value proposition. Emails are kept under 120 words with a single clear call to action.
This project is intentionally lean. You need only 5 MCP servers to run the full pipeline:
| MCP Server | Agent | Purpose |
|---|---|---|
| apolloio-mcp | Lead Researcher | Company search, firmographics, tech stack |
| crunchbase-mcp | Lead Researcher | Funding data, investors, company profiles |
| hunter-mcp | Contact Enricher | Email discovery and verification |
| linkedin-mcp | Contact Enricher | Professional profiles, job titles |
| exa-ai-mcp | Outreach Strategist | Company news, signals, personalization data |
All hosted on the Vinkius AI Gateway. Deploy each one in under a minute and get your SSE endpoint URL.
The Vinkius AI Gateway offers additional sales intelligence MCP servers you can add with a one-line config change:
Browse the full catalog of 2,600+ production-ready MCP servers at vinkius.com/en/categories.
git clone https://github.com/vinkius-labs/crewai-mcp-lead-gen-agents.git
cd crewai-mcp-lead-gen-agents
python -m venv .venv
source .venv/bin/activate # Linux/macOS
# .venv\Scripts\activate # Windows
pip install -e .
cp .env.example .env
Add your Gemini API key and 5 Vinkius MCP URLs.
# Validate configuration
lead-gen validate
# Generate leads for FinTech VP of Engineering
lead-gen prospect "FinTech" "VP of Engineering" \
--size "50-500" \
--product "API monitoring platform" \
--value "reduce API downtime by 90%"
# Generate leads for SaaS Head of Product
lead-gen prospect "SaaS" "Head of Product" \
--size "100-1000" \
--product "user analytics platform" \
--value "increase feature adoption by 3x"
The system produces a markdown report containing:
| Section | Content |
|---|---|
| Company Profiles | 10 qualified companies with firmographics, fit scores, and reasoning |
| Decision-Maker Contacts | Verified emails, LinkedIn profiles, titles, seniority |
| Outreach Templates | 10 personalized emails with subject lines and CTAs |
| Summary Metrics | Total leads, average fit score, email coverage rate |
@CrewBase decoratorsCompanyProfile, ContactProfile, OutreachMessagemcps= field with SSE transport to Vinkius AI GatewayThe Model Context Protocol is an open standard for connecting AI systems to external tools. See modelcontextprotocol.io.
Yes. Add the HubSpot, Salesforce, or Pipedrive MCP server and extend the outreach agent to push leads directly.
This tool discovers publicly available business contact information. However, compliance depends on your jurisdiction and how you use the data. Consult your legal team before using enriched contact data for outbound campaigns.
Please read the Contributing Guide before submitting a pull request.
MIT — see LICENSE.
Built by Vinkius Labs with CrewAI and the Vinkius AI Gateway.
praniketkw /
6 AI agents collaborate to build complete web applications in minutes. CrewAI-powered software development team that generates FastAPI backends, responsive frontends, tests, and deployment configs - demonstrating 95% automation with human oversight for the final 5%.
Shindevrp /
Learning CrewAI is a Python project that demonstrates the CrewAI framework for building AI-powered agent crews
40/100 healthvinkius-labs /
AI agent crew that generates institutional-grade crypto market intelligence reports with real-time data from CoinGecko, Binance, SEC EDGAR, FRED and more using CrewAI + 15 MCP servers + Google Gemini.
66/100 healthThis project demonstrates how to build multi-agent AI systems using CrewAI and Python with a local Mistral-7B Instruct (GGUF) model. It includes agent collaboration for tasks like competitor analysis and content strategy generation, running fully offline using llama-cpp-python.
42/100 health