Open source · Public
GTM Skills
Before the full chain, the writing and judgment layer shipped on its own: seven skills that turn ten years of cold-email method into something a coding agent applies consistently. Install with npx skills add Zevenue/gtm-skills.
What was actually hard
Every team rediscovers the same outbound rules the hard way: which signals are worth a sentence, how to score them, what a three-email sequence looks like when it is written to be delivered and answered rather than admired.
In order
- Context. an interview that writes the offer and ICP files every other skill reads, once per workspace.
- Signal builder. rank a prospect's signals 1 to 10 with provenance and pick the campaign approach: pain-led, value-led, or segment fallback.
- Creative variable. spec the non-obvious personalization variables: names, grammar forms, sources, extraction prompts, fallbacks.
- Email writer. three-email sequences on the Situation, Insight, Inquisition method, with a QA checklist enforced.
- Prospect posts. scan recent LinkedIn posts for a theme before outreach.
Decisions that made it work
- Methodology belongs in files the model reads, not prompts a person remembers.The skills are the same ones Zevenue applies on client work; publishing them forced the method to be explicit enough to survive without me.
Outcomes
Public and installable. Folded into Headless GTM as its writing layer.
Why it matters
Ten years of cold email, written down precisely enough for an agent to apply.