nous
1.2kGet started
← Skill Library
GTM

/lookalike-builder

Paste up to five companies you would love to work with, or describe the niche, and a similarity model finds the lookalikes. It narrows them with keywords and firmographics, scores every company against your ICP, and saves the company table into your context graph.

aiark aiark claude claude nous nous
Install it View source ↗
SEEDS

How it works

The company table is the deliverable. Four steps, and the search and the count are free to preview before a single credit moves.

  1. Discover the lookalikes

    Paste up to five seed companies, or describe the niche. The similarity model returns every company that looks like them, and the whole search is free to preview.

    aiark aiark

  2. Narrow and exclude

    Tighten the set with include and exclude keywords, headcount and location. Companies already in your graph are excluded automatically, so you never find the same one twice.

    aiark aiark

  3. Score against your ICP

    Every company is scored zero to one hundred with a reason attached. You see the count and the sample, then you approve, and only then does anything get spent.

    claude claude

  4. Save the company table

    The scored companies land in Nous as a list. Hand that list to company-people and each company becomes a decision maker.

    nous nous

Make it your own

Every skill is a plain markdown file in your own folder. Read it, change it, and it is yours.

Refine it to your ICP

The scoring reads your GTM profile from Nous. Sharpen icp.md, sync it, and the next run ranks the lookalikes against the definition you actually sell to.

Make it yours

Open the SKILL.md and change what it does. Add your own exclude vocabulary, move the headcount band, or push the stop point further down the pipeline.

Wire it into your workflow

By default it stops at the company table, because the LinkedIn people layer is more complete on small teams. Hand off to company-people, or keep going inside this skill on companies of fifty or more.

See it run

Five seeds, four hundred lookalikes, a scored company table in about three minutes.

walkthrough · coming soon

Install it

Paste it into your terminal. Then use /lookalike-builder inside Claude Code.

terminal 
curl -sL https://raw.githubusercontent.com/NousC/gtm-skills/main/.claude/skills/lookalike-builder/SKILL.md --create-dirs -o ~/.claude/skills/lookalike-builder/SKILL.md

Then run it

Open Claude Code and type /lookalike-builder. Paste your seeds, or describe the niche. It builds the spec, shows you the count and a sample, and waits for you before it saves.