enterprise-gtm-skills

Enterprise GTM Skills

Ten Agent Skills for enterprise and regulated-B2B product marketing. Written by a product marketing manager at SAP in Berlin, from work in the motion these skills describe.

npx skills add rushogen/enterprise-gtm-skills

Why this exists

Every general marketing skills collection assumes a product-led, self-serve motion. The vocabulary gives it away: paywalls, popups, signup flows, activation funnels, freemium churn, trial-to-paid conversion. All of it is useful, and none of it applies when the deal takes nine months, four functions inside the buyer hold a veto, and the person you convinced does not have signing authority.

Enterprise product marketing is a different job:

• The buyer is a committee with veto holders. InfoSec, privacy, legal, procurement, enterprise architecture and — in Germany — the works council can each stop a deal independently, and none of them care about your value proposition.
• Marketing's customer is the field, not the buyer. A launch succeeds or fails on whether sellers pitch it, which depends on quota retirement, a SKU in CPQ, and a demo environment — not on the announcement.
• Nothing is measurable inside a quarter. Revenue from a launch arrives two sales cycles later. If you cannot name the leading indicators, you cannot manage the programme.
• Third parties gate the market. A Magic Quadrant dot or a Wave score decides which vendors reach a shortlist, and both are decided eighteen months before publication.
• Compliance is a marketing surface. "Bank-grade encryption" becomes a question in a vendor risk assessment. Article 50 of the EU AI Act now governs what your chatbot has to say in its first message.
• The price is not the offer. List price is an anchor and a governance instrument. Every deal is negotiated.

None of that is a variation on self-serve marketing. It is a different set of artefacts, stakeholders, gates and failure modes, and that is what these skills contain.

The skills

Each skill is a SKILL.md under 500 lines, with deeper material in references/ and templates in assets/, loaded on demand.

What this does not cover

An honest scope statement, because a pack that claims to cover everything is usually written by someone who has done none of it.

Not covered, deliberately:

• Product-led growth, self-serve funnels, trials, freemium, paywalls, onboarding, activation and in-product growth loops. There are good collections for that; this is not one, and mixing the two produces advice that is wrong in both directions.
• Demand generation mechanics: paid media buying, SEO, email deliverability, landing page conversion optimisation, marketing automation configuration.
• Brand, visual identity and creative production.
• Sales methodology. These skills assume MEDDIC, Challenger or whatever your organisation runs, and produce the marketing artefacts that support it.
• Marketing operations tooling and attribution modelling.
• Public sector procurement in depth. It is named where it constrains a decision, but framework bidding is its own discipline.
• Non-European regulatory regimes. The regulatory content is EU and DACH, which is where the author works.

Covered but thin, and honestly labelled:

• Analyst firms other than Gartner and Forrester. IDC, ISG, Everest and HFS are named with their contexts, not detailed.
• APJ and North American market entry. The country-sequencing skill is European.
• Vertical-specific motion — financial services, healthcare, public sector, and industrial — is referenced where it changes a gate, not developed.

Two standing caveats. Nothing here is legal advice; the regulatory content tells you which questions to take to counsel. And every date and regulatory position is current to August 2026 and should be verified before you rely on it.

Installing

Via the skills CLI:

npx skills add rushogen/enterprise-gtm-skills

Manually, for Claude Code — copy the skill directories into ~/.claude/skills/ for personal use or .claude/skills/ in a project:

git clone https://github.com/rushogen/enterprise-gtm-skills.git
cp -r enterprise-gtm-skills/skills/* ~/.claude/skills/

The pack also ships .claude-plugin/plugin.json, so it can be installed as a plugin from a marketplace entry.

Skills are portable. The frontmatter uses only the six fields defined by the Agent Skills specification — name, description, license, compatibility, metadata, allowed-tools — so they load on claude.ai, through the Skills API, and in the other conforming clients without modification. There is no top-level version field in the spec; versioning lives under metadata.version.

Validating

python3 scripts/validate_skills.py .

Checks name-to-directory agreement, the portable-field restriction, length limits, and that every relative link resolves. Dependency-free.

python3 scripts/run_evals.py
python3 scripts/run_evals.py --list enterprise-competitive

Twenty eval cases across four skills, with a grading harness. See evals/README.md, including its limitations section.

Using them well

These skills produce better output when you give them the situation rather than the task. "Write a battlecard" gets a template. "We lose to this incumbent at the security review stage in German financial services and I do not know why" gets the diagnosis, and the diagnosis is where the value is.

Several skills will push back on the question as asked — the analyst relations skill will not draft a Magic Quadrant placement appeal, and the competitive skill will not invent traps without a verifiable weakness. That is deliberate, and it is tested.

Contributing

See CONTRIBUTING.md. The bar is specificity: a real artefact name, a real stage gate, a real failure mode you have watched happen. Corrections from people who have run these programmes are the most valuable contributions here, particularly on regulatory content and on markets outside Europe.

Licence

MIT. Copyright (c) 2026 Ruslan Shogenov. Vendor them into an internal enablement repository without a legal review.