Frontal GTM OS

Frontal's Go-To-Market Engineering workspace. Clone this repo, run the onboarding script, and use Claude Code to run go-to-market delivery: building TAM lists, scoring companies, writing campaign copy, enriching contacts, loading campaigns, and analyzing performance.

Part of the Frontal GTM resource hub.

Prerequisites

• Claude Code installed (npm install -g @anthropic-ai/claude-code)
• Git
• Python 3.9+

Onboarding (first time)

Run the onboarding script once on your machine:

# 1. Clone the repo
git clone https://github.com/Frontal-so/gtm-os.git
cd gtm-os

# 2. Copy the env template and fill in your own keys
cp .env.example .env
# edit .env — see "API Keys" in SETUP.md

# 3. Run onboarding
./scripts/gtme-onboard

The onboarding script will save your profile locally, collect your keys, add gtm-work to your PATH, and launch Claude Code to walk you through each workspace.

Daily workflow

# Open a new terminal and type:
gtm-work

This checks prerequisites, pulls the latest changes, loads your context into Claude Code, and asks which client you want to work on.

Slash commands

Command	What it does
/client <name>	Load full context for a specific client
/status	Show current client status, live stats, and next steps
/save	Commit all work and log the session
/load <csv-path>	Load a CSV into your data layer (auto-maps columns, dry-run first)
/report <name>	Pull last 7 days campaign performance for a client
/onboard	Re-run guided onboarding for a client (API keys, TALs, ICP)

Skills

14 specialized skills the assistant loads on demand:

Skill	What it helps with
Apollo Search	Company/contact search with 65+ filters
Lima Data API	LinkedIn discovery, email finding, batch processing
Clay Mastery	Table architecture, email waterfall, Claygent
Contact Enrichment	3-layer enrichment (Linkup, Lima, Clay)
List Dedup	LinkedIn URL dedup, company name dedup
Crawford Method	PVP/PQS cold email framework
Cold Email Copy	4-part email anatomy, best-in-class templates
Campaign Delivery	9-phase build checklist
Instantly API	Campaign creation, lead loading, analytics
TAM Scoring	Point-based scoring models, tier thresholds
Signal Detection	Intent signals with freshness rules
Fireflies Usage	Meeting transcript search, action items
Meta Ads Scraper	Company discovery via the Meta Ad Library

Repo structure

gtm-os/
├── CLAUDE.md                  # AI assistant instructions (auto-loaded)
├── README.md                  # This file
├── SETUP.md                   # API keys and setup detail
├── schema.sql                 # Database schema
├── schema-lists.sql           # List management schema
├── schema-intelligence.sql    # Intelligence schema
│
├── clients/                   # One folder per client
│   ├── _template/             # Copy for new client onboarding
│   └── example-client/        # Blank example (context.md)
│
├── skills/                    # 14 reusable GTM workflow skills
│   └── README.md              # Full skill index
│
├── scripts/                   # Session launcher, onboarding, data helpers
└── .claude/commands/          # Slash commands

Data architecture

• This repo: context docs, scoring scripts, campaign copy, skill files, workflow templates
• Your data layer (Supabase): TAM/TAL data, contact records, campaign performance, API keys, session logs
• Sending platforms: Instantly, Lemlist, Smartlead (via API keys you provide)

Working rules

• READ-ONLY on all external services unless you explicitly confirm a write
• Use lowercase-kebab-case for all file and folder names
• Prefix reports with dates: 2026-03-19-campaign-report.md
• Always run /save before ending a session

Credit

Built and maintained by Frontal.

License

MIT, see LICENSE.