gtm-setup
Skill16 starsChanged 49 days agoApache-2.0View on GitHub
First-run setup for the GTM pipeline. Asks a few questions, reads the user's website, and writes context/icp.md, personas.md, segments.md, exclusions.md and do-not-contact.csv, copies gtm.config.yaml and .env from their examples, then runs scripts/doctor.py and suggests a first brief. Use when the user says "/gtm-setup", "set this up", "get started", "onboard me", or when /gtm finds the context files missing or still templates.
- Reads credentials
1 other project uses a version of this.
Adopted by
Most recently taken up by
- cognyai/claude-code-marketing-skillstheir own version, first seen Oct 12, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- cognyai/claude-code-marketing-skills★ 106same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- activateSkill
- company-discoverySkill
- company-enricherSkill
- contact-qualifierSkill
- contact-sourcerSkill
- email-finderSkill
- gtmSkill
- phone-finderSkill
The file
Files
Set up this checkout so /gtm <brief> works. The user should be able to finish in about five
minutes without reading the docs. Input, if any: $ARGUMENTS
Rules
- Never overwrite an existing, filled-in file. If
context/icp.mdexists without<placeholder>text, show it and ask what to change instead. - Never write an API key. Copy
.env.exampleto.envonly if.envis missing, and tell the user which keys to fill in. Keys never go in any other file. - Never invent a customer, a competitor or a seed company. Seeds and do-not-contact entries come from the user. Titles and segments you infer are drafts, and you say so.
Steps
- Copy what is missing (never overwrite):
[ -f gtm.config.yaml ] || cp gtm.config.example.yaml gtm.config.yamland the same for.env. - Ask, in one message (accept short answers; skip what the user doesn't know):
- Your website URL.
- What you sell, in one sentence.
- Who buys it: job titles, and who else is involved in the decision.
- Where: countries or regions.
- Three to ten of your best customers (names or domains). These become look-alike seeds.
- Anyone you must never contact: current customers, partners, opt-outs. A list of domains/emails, a CSV path, or "none for now".
- Company types you never sell to (e.g. agencies, competitors).
- Read the website (home, product, customers/case-studies, pricing pages; WebFetch) for the product, the buyer, and the industries it names. Use it to draft; do not quote marketing claims as facts about the market.
- Draft the files from the
.exampletemplates, keeping their headings so the pipeline can read them:context/icp.md— product, segments in priority order, default geography, default seniorities, the seeds from answer 5, anti-fit signals.context/personas.md— one block per buyer role from answer 3, each withTitles:,Also-known-as:(sector and local-language variants for the user's region), a vertical-appropriateSeniority:,Why relevant:and aDisambiguation:line naming the wrong senses of the title.context/segments.md— A/B/C criteria from the segments, the default 0–10 rubric and thresholds.context/exclusions.md— answer 7 plus the standard always-skip titles.context/do-not-contact.csv— headeremail,domain,phone,linkedin_url,reason, then answer 6, one entry per row with a reason. If the user said "none", write the header only and tell them that means "checked against an empty list".
- Show the drafts (personas and segments in full, the rest summarized) and ask for corrections. Apply them, then write the files.
- Check the setup:
python3 scripts/doctor.py. Explain each FAIL and warning in one line with its fix. The usual remaining item is provider keys: say which one key would give the biggest improvement for this user's stages (seedocs/single-provider.md), and that the pipeline runs without any keys on the builtin web research, just thinner. - Suggest a first brief built from their answers, e.g.
/gtm find <persona> at <segment> in <region> for <product>, and mention/gtm status <id>and/gtm resume <id>.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

