gtm-wizard 0.2.1

MCP server packaging Go-To-Market Engineering expertise for AI agents

GTM Wizard

The Senior GTM Engineer in Your Pocket - Go-To-Market Engineering expertise for AI agents via MCP.

An MCP (Model Context Protocol) server that brings world-class GTM Engineering expertise into AI-powered workflows. Built from real production systems handling high-velocity lead pipelines.

Note: GTM = Go-To-Market, not Google Tag Manager.

What is GTM Wizard?

GTM Wizard is the foundation layer for Agentic GTM - AI agents that can build and operate GTM machines.

GTM Wizard IS	GTM Wizard is NOT
An expertise layer AI agents USE	A tutorial or teaching tool
Action-oriented tools with structured outputs	A collection of templates to copy
Flexible components for different contexts	Educational content explaining concepts
The "GTM brain" for autonomous operations	A replacement for strategic thinking

Features

Tools (6 available)

Tool	Purpose
score_lead	Calculate lead scores with transparent breakdown
classify_role	Classify job titles into decision-making tiers
classify_industry	Determine industry fit based on your ICP
determine_routing	Route leads to appropriate engagement tracks
check_disqualification	Check for disqualifying factors
diagnose_rate_limiting	Debug API rate limit issues

Prompts (4 available)

Prompt	Output
lead-qualification-workflow	Structured qualification result with routing decision
icp-definition	YAML ICP configuration for lead tools
outbound-campaign-design	Campaign blueprint with sequence and metrics
lead-scoring-calibration	Scoring model calibration recommendations

Resources (5 available)

GTM Engineering knowledge accessible via gtm://foundations/{resource-id}:

• what-is-gtm-engineering - Core definitions and skills
• gtm-archetypes - 6 specialization types
• context-factors - 8 factors that shape decisions
• principles-not-recipes - GTM Wizard philosophy
• knowledge-taxonomy - Full domain map

Installation

From PyPI (Recommended)

pip install gtm-wizard

From Source (for contributors)

git clone https://github.com/MathewJoseph1993/gtm-wizard.git
cd gtm-wizard
pip install -e ".[dev]"

Quick Start

Claude Desktop

Add to your Claude Desktop config:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json Linux: ~/.config/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "gtm-wizard": {
      "command": "uvx",
      "args": ["gtm-wizard"]
    }
  }
}

Alternative (if you installed with pip):

{
  "mcpServers": {
    "gtm-wizard": {
      "command": "python",
      "args": ["-m", "gtm_wizard.server"]
    }
  }
}

Cursor

Add to Cursor MCP settings (Settings → MCP → Add Server):

{
  "mcpServers": {
    "gtm-wizard": {
      "command": "uvx",
      "args": ["gtm-wizard"]
    }
  }
}

Claude Code CLI

claude mcp add gtm-wizard -- uvx gtm-wizard

Or if installed with pip:

claude mcp add gtm-wizard -- python -m gtm_wizard.server

Example Usage

Qualify a Lead

Qualify this lead: [e-mail], VP of Marketing at Acme Corp

GTM Wizard will run the full qualification pipeline and return:

QUALIFICATION RESULT
====================
Lead: [e-mail]
Status: QUALIFIED
Score: 65/95 (68%)
Tier: B
Routing: medium_touch

Key Factors:
- Role: VP Level - 25 points
- Industry: Unclassified - 10 points
- Company Size: 20 points

Recommended Action: Enroll in nurture sequence with sales oversight

Design a Campaign

Design an outbound campaign targeting VP of Sales at SaaS companies, goal is booking demos

Returns a complete CAMPAIGN_BLUEPRINT with targeting, sequence, messaging framework, and metrics.

Build ICP Configuration

Help me build an ICP config for my B2B SaaS product

Returns a structured ICP_CONFIG in YAML that feeds into lead scoring tools.

Development

# Sync dependencies (UV)
uv sync --all-extras

# Run all checks
make all

# Individual commands
make test        # Run tests (50 tests, 85% coverage)
make lint        # Lint code
make format      # Format code
make type-check  # Type checking (strict)
make serve       # Run the MCP server
make inspect     # Open MCP Inspector

Testing with MCP Inspector

make inspect

Opens browser at http://localhost:6274 for visual tool testing.

Roadmap

See ROADMAP.md for full details.

Current: v0.2 - Lead Intelligence (Complete) Next: v0.3 - Outbound Automation

Future milestones include:

• Outbound campaign execution tools
• Data & analytics capabilities
• CRM integration patterns
• Agentic integrations (HubSpot MCP, email tools, context awareness)

Contributing

Contributions welcome! See CONTRIBUTING.md for guidelines.

Content Guidelines:

• Action-oriented - tools DO things, don't explain things
• Structured outputs - produce configs, blueprints, decisions
• Integration-ready - design for CRM/tool connections
• Expert application - apply GTM expertise, don't lecture

License

MIT License - see LICENSE for details.

---

Built by Mathew Joseph - GTM Engineer