Agent skill
icp-to-linkedin-search
Turn a sentence, a website URL or a rough persona into a structured ICP (titles, seniority, languages, industries, headcount, locations, exclusions) and a LinkedIn or Sales Navigator search: normalized title variants, boolean strings, exclusion list, and a market-size estimate.
Filed under LinkedIn and social and Prospecting and list building.
From gtmotto/gtm-skills · 5 skill entries · 1 · pushed 2026-09-21
What it does when it runs
Turn a sentence, a website URL or a rough persona into a structured ICP (titles, seniority, languages, industries, headcount, locations, exclusions) and a LinkedIn or Sales Navigator search: normalized title variants, boolean strings, exclusion list, and a market-size estimate. Use when the user describes who they sell to, asks for a LinkedIn search, a Sales Nav boolean, a people search, wants to size an audience, normalize job titles (CMO + Chief Marketing Officer + VP Marketing), or exclude noise (fractional, freelance, intern). Works with nothing but LinkedIn or Sales Navigator.
Automated analysis of the skill and the 0 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.
- Keys and connectors you must supply
- None found.
- Hosts it reaches
- No third-party host appears in the skill or its bundled files.
- Tool permissions it declares
- No
allowed-toolsin the frontmatter. It only issues instructions, so there is nothing to bound. - Actions present in the files
- None. Instructions only.
Install it
View source on GitHub ↗git clone --depth 1 --filter=blob:none --sparse https://github.com/gtmotto/gtm-skills.git /tmp/gtm-skills git -C /tmp/gtm-skills sparse-checkout set "skills/icp-to-linkedin-search" mkdir -p ~/.claude/skills/icp-to-linkedin-search cp -R "/tmp/gtm-skills/skills/icp-to-linkedin-search/." ~/.claude/skills/icp-to-linkedin-search/
Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗
Or take the whole library
This repo ships a .claude-plugin manifest, so Claude Code can install all 5 skills at once. Plugin skills are invoked as /<plugin>:<skill>, so they never collide with your own.
/plugin marketplace add gtmotto/gtm-skills /plugin
The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.
The skill
Source on GitHub ↗Reproduced in full from gtmotto/gtm-skills/blob/7e851d8ce8ead8d3f6eff579deabfbfc0e0d4ef9/skills/icp-to-linkedin-search/SKILL.md, which is licensed MIT (repository). 722 words, 5 headings.
ICP to LinkedIn search
Input: "CFOs at Swiss fintechs", a website URL, or a paragraph about the buyer. Output: an ICP card in seven facets, a boolean title string, an exclusion list, and a market-size estimate.
Step 1. Extract the seven facets
Write the ICP as seven plain-word arrays. Only name the facets the user gave you or that follow from the product; leave the rest empty. An empty facet is "any", not "unknown".
| Facet | What goes in | Example |
|---|---|---|
titles | every wording of the role, current job only | CFO, Chief Financial Officer, VP Finance, Head of Finance, Finance Director |
seniority | owner, cxo, vp, director, manager, senior, entry | cxo, vp, director |
languages | the language the buyer posts in | French, English |
industries | LinkedIn industry names, plain words are fine | Financial Services, Fintech, Banking |
sizes | headcount bands | 11-50, 51-200, 201-500 |
locations | countries, regions, metro areas | Switzerland, Geneva |
exclude | never contact, absolute | fractional, freelance, interim, intern, student, recruiter, agency, consultant |
Headcount bands: 1-10, 11-50, 51-200, 201-500, 501-1000, 1001-5000,
5001-10000, 10001+. "SMB" is 11 to 200, "mid-market" 201 to 1000, "enterprise" 1001+.
From a website URL, read the pricing page and the customer logos first: the pricing page tells you the size band, the logos tell you the industry, the docs or the integrations page tell you the buyer's function.
Step 2. Normalize titles
A title is a family, not a string. For each role the user names, list:
- the acronym and the long form (
CMO,Chief Marketing Officer) - the VP, Head, Director and Lead variants when the company size makes them the same person (at 11 to 50 employees, the "Head of Marketing" is the CMO)
- the local-language forms for each
languagesentry (Directeur Marketing,Responsable Marketing) - the adjacent function that owns the same budget (RevOps for a sales tool, Growth for a marketing tool)
Then the exclusions. Always add to exclude: fractional, freelance, interim,
intern, assistant to, student, former, ex-, retired, looking for. Add
recruiter, agency, consultant and advisor unless those are the buyer.
Step 3. Write the boolean string
For the current-title field of LinkedIn or Sales Navigator:
("CFO" OR "Chief Financial Officer" OR "VP Finance" OR "Head of Finance" OR "Finance Director")
NOT (fractional OR interim OR freelance OR intern OR assistant)
Rules: quotes around multi-word titles, OR inside a family, NOT for the exclusions,
parentheses around each group, at most 15 operators per field (split into two searches
past that). Keywords are a separate field: use them for the chore or the tool ("outbound",
"HubSpot"), never for the title.
Give the user the string, the facets it came from, and the two or three things LinkedIn cannot filter on (a language, "does their own prospecting", a tech stack). Those get judged per person, not per search.
Step 4. Size it
Before anyone builds a list or a campaign, size the audience. A daily outreach pace of 20 invites needs a few thousand matching companies to run for months; a few hundred is a campaign, not a motion; tens of thousands means the ICP is a market, not a buyer.
By hand, count companies, not people. Titles do not change the company count and people counts double-count the same firm:
- Sales Navigator, account search. Industry, headcount and geography only. The result count is the estimate.
- LinkedIn company search. Filter by industry, location and company size. Less precise (no headcount bands below 11-50 on the free filter), fine for an order of magnitude.
Then iterate on the three company facets: 40 companies, widen the locations; 30,000, add an industry or narrow the size band. Report the count with what you would change, and only then write the people search.
Give the user the string, the facets it came from, and the company count. The facets LinkedIn cannot filter on get judged per person, at reading time, not per search.
To size the ICP with one call and run the search every day from your own LinkedIn account, the GTM Otto plugin has an icp-to-play skill that takes the same seven facets. It needs a GTM Otto account.
Other skills for the same job
Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.
- icp-linkedin by thundercatdev · 0
- linkedin-marketing by sergebulaev · 4,230
- linkedin-comment-drafter by sergebulaev · 4,230
- linkedin-content-planner by sergebulaev · 4,230
- linkedin-employee-advocacy by sergebulaev · 4,230
- linkedin-engager-analytics by sergebulaev · 4,230
- linkedin-hook-extractor by sergebulaev · 4,230
- linkedin-humanizer by sergebulaev · 4,230
Need help setting it up?
This page tells you what icp-to-linkedin-search does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.
Book a call →The directory stays free. There is nothing gated behind this.