The GTM Operator Brainstorming Skill: An Outbound Strategy Framework for Claude

This guide covers how to install the skill, how to use it, and a set of starter prompts to get you running immediately.

---

01What this skill is

A brainstorming partner for GTM operators. It compiles the frameworks, principles, and decision logic from five operators running cold outbound at scale. You bring the context about your business and your current campaign. The skill helps you think through the right approach before you start building.

This skill does not build campaigns, write copy, or pull lists. It diagnoses. It surfaces frameworks. It helps you find the right angle and the right questions before you touch any tooling.

---

02Installation

Step 1: Save the two skill files

You need both files for the skill to work properly.

Both files are included in the PDF download.

• SKILL.md: the principles and diagnostic layer
• frameworks-library.md: the deeper stage-by-stage framework reference

Step 2: Create a new Claude Project

1. Open Claude at claude.ai
2. Click Projects in the left sidebar
3. Click New Project
4. Name it something like "GTM Operator Brainstorm"

Step 3: Upload both files to the Project

1. Inside your new Project, click Add content or the upload icon
2. Upload SKILL.md
3. Upload frameworks-library.md
4. Both files should now appear in your Project's knowledge base

Step 4: Start a conversation

Open a new chat inside the Project. The skill is now active. Claude will reference both files when you ask GTM and outbound questions.

---

03How to use it

The skill works best when you give it context before asking for advice. The more specific you are about your situation, the sharper the diagnosis.

The basic pattern:

1. Tell it where you are (stack, volume, what you're running)
2. Tell it where you're stuck
3. Let it ask you a couple of diagnostic questions
4. Work through the framework it surfaces

Do not ask it to build anything. If you want to move from brainstorm to build, take the framework it gives you and use that to brief your Claude Code workflow or your team.

---

04Starter prompts

These are ready to copy and paste. Edit the brackets for your situation.

---

Prompt 1: Full diagnostic

I'm building an outbound system for [describe your business and ICP].
Right now I'm using [Clay / Claude Code / both / neither].
I'm sending roughly [X emails per month].
The thing I'm most stuck on is [describe the bottleneck].
Run me through the diagnostic and tell me where the real problem is.

---

Prompt 2: Stack decision

I need to decide how to split work between Clay and Claude Code
for my outbound. My volume is [X contacts per month].
My team is [technical / non-technical].
I need to be able to [see the logic visually / just have it run automatically].
Help me think through the right split.

---

Prompt 3: Enrichment waterfall

Walk me through how to structure an email enrichment waterfall.
I'm currently using [list your current providers or say "nothing yet"].
My target audience is [describe ICP].
I need to find emails for roughly [X contacts per month].
Tell me the right shape for this and where my current setup has gaps.

---

Prompt 4: List building and signal

I'm building a list of [describe target companies and titles].
Right now I'm pulling from [Apollo / Clay / CSV / nothing structured].
I don't have a signal layer yet.
Help me think through what signals actually matter for this ICP
and how to layer them on top of my list.

---

Prompt 5: Scoring and segmentation

I want to set up a tiering system for my outbound list.
My best clients look like [describe closed-won profile].
I'm targeting [describe ICP].
Walk me through how to build a scoring framework
and how to map each tier to a different outreach approach.

---

Prompt 6: Copy and feedback loop

My reply rates have dropped on [describe campaign type].
I've been running the same copy framework for [X months].
I have [describe what analytics you have access to].
Help me diagnose whether this is a copy problem, a list problem,
a signal problem, or something else.

---

Prompt 7: Reinforcement loop

Every time we launch a new campaign it feels like we're starting from scratch.
We're not capturing what worked.
We have [describe what data you have: reply data / meeting data / call recordings / none].
Help me think through how to build a feedback loop
so each campaign starts smarter than the last.

---

Prompt 8: Full pipeline audit

I want a full audit of my current outbound pipeline.
Here's how it works right now: [describe your current process step by step].
Tell me which stage has the biggest gap
and what the most important thing to fix first is.

---

05Tips for getting the best output

Give volume numbers when you have them. The skill surfaces different frameworks depending on whether you're at 500 emails a month or 500,000.

Tell it your stack. Clay-heavy, Claude Code-heavy, or neither changes the recommendations significantly.

Be honest about what's not working. The skill is designed to find the real bottleneck, which is often one stage upstream from where you think the problem is.

Use it before you build. The skill's job is to get you to the right shape before you touch any tooling. If you build first and brainstorm second, you'll rebuild.

---

06What to do after the brainstorm

Once the skill has given you a diagnosis and a framework:

1. Take the framework and use it to write a brief for your Claude Code workflow
2. Share it with your team as context before you start the build
3. Use it to write your CLAUDE.md if you're setting up a new Claude Code project
4. Come back after your first campaign run and use the reinforcement loop prompts to close the feedback loop

---