account-research-brief

The GTM Engineer's Playbook for the Claude Code Era

10 installable Claude Code skills that turn your terminal into a GTM command center.

Built by Earleads — GTM Engineering as a Service.

---

What This Is

A collection of Claude Code skills designed for GTM engineers, sales operators, and revenue teams. Each skill encodes a repeatable methodology — not a one-shot prompt. Install them, trigger them with natural language, and they produce structured output you can use immediately.

These skills are the building blocks of a modern GTM operating system. They chain together: your ICP feeds your Signal Scanner, which feeds your Qualification Scorer, which feeds your Outreach Sequences.

The Skills

Core Skills (7)

Bonus Skills (3)

How They Chain Together

ICP Architect ─────────────┐
  "build my ICP"           │
  produces: docs/icp.md    │
                           ▼
Signal Scanner ◄─── reads icp.md for scoring criteria
  "scan for signals"       │
  produces: docs/signals/  │
                           ▼
Qualification Scorer ◄──── reads icp.md + signals/
  "score this lead"        │
  produces: docs/pipeline/ │
                           ├──────────────────────┐
                           ▼                      ▼
Outreach Sequence Builder  Meeting Prep Brief
  reads icp.md + accounts/ reads accounts/ + signals/
  produces: docs/sequences/ produces: docs/meeting-prep/
                           │
                           ▼
Account Research Brief ◄── standalone or feeds into above
  "research [company]"
  produces: docs/accounts/

Competitive Battlecard ◄── standalone reference
  "battlecard for [X]"
  produces: docs/battlecards/

Weekly GTM Report ◄──────── reads ALL docs/ subdirectories
  "weekly report"
  produces: docs/reports/

Installation

Option 1: Quick Install (recommended)

git clone https://github.com/Othmane-Khadri/gtm-engineer-playbook.git
cd gtm-engineer-playbook
./install.sh

Option 2: Manual

git clone https://github.com/Othmane-Khadri/gtm-engineer-playbook.git
cp -r gtm-engineer-playbook/.claude/skills/gtm-playbook/ your-project/.claude/skills/gtm-playbook/

Option 3: Cherry-pick specific skills

# Only install the skills you want
cp -r gtm-engineer-playbook/.claude/skills/gtm-playbook/icp-architect/ your-project/.claude/skills/gtm-playbook/icp-architect/
cp -r gtm-engineer-playbook/.claude/skills/gtm-playbook/signal-scanner/ your-project/.claude/skills/gtm-playbook/signal-scanner/

Uninstall

rm -rf .claude/skills/gtm-playbook/

Prerequisites

• Claude Code installed and authenticated
• A Claude Code project (any project — these skills work anywhere)
• WebSearch access (most skills use it for real-time intelligence)

Recommended MCP Servers

These skills work standalone, but they're even more powerful with MCP servers connected:

Quick Start

After installation, open Claude Code in any project and try:

1. "Build my ICP" → walks you through defining your target customer
2. "Research Acme Corp" → produces a 1-page intelligence brief
3. "Scan Acme Corp for buying signals" → finds and scores live signals
4. "Score my leads" → qualifies your pipeline with tier classification
5. "Build an outreach sequence for post-fundraise" → ready-to-send messages

Output Structure

All skills write to a docs/ directory in your project root:

docs/
├── icp.md                          # ICP Architect output
├── signals/
│   ├── README.md                   # Signal summary
│   └── {company-name}.md           # Per-company signals
├── accounts/
│   └── {company-name}.md           # Account research briefs
├── pipeline/
│   └── scored-leads.md             # Qualification scores
├── sequences/
│   ├── README.md                   # Sequence index
│   └── {trigger-name}.md           # Per-trigger sequences
├── meeting-prep/
│   └── {company}-{date}.md         # Meeting briefs
├── battlecards/
│   ├── README.md                   # Competitor index
│   └── {competitor-name}.md        # Per-competitor battlecards
├── reports/
│   ├── README.md                   # Report index
│   └── week-{YYYY-WNN}.md         # Weekly reports
├── crm-hygiene-report.md           # CRM audit
└── agents/
    └── {agent-name}/               # Agent architecture plans
        ├── architecture.md
        ├── run.sh
        ├── prompt.md
        ├── config.yaml
        └── README.md

Complementary Resources

These skills pair well with existing GTM resources:

Philosophy

Skills > Prompts. A prompt gives you a one-time answer. A skill encodes a methodology that produces consistent, high-quality output every time. The hard part of GTM engineering isn't writing the prompt — it's knowing the right questions to ask, the right frameworks to apply, and the right sequence to follow.

Chain, don't silo. Each skill reads the output of other skills. Your ICP informs your signal scoring. Your signals inform your outreach. Your outreach results feed your weekly report. The whole system compounds.

Shell orchestrates, Claude thinks. For autonomous agents, let the shell handle infrastructure (API calls, scheduling, file management) and let Claude handle intelligence (scoring, writing, deciding). This separation keeps both layers simple and debuggable.

---

Built by Earleads