Build Campaign Context & Hypotheses4 skills
Foundation layer. Build and maintain your company context file and generate testable pain hypotheses.
Sales Blueprint
Safety checked21 GTM skills for research, list building, enrichment, segmentation, email generation, and sending. Use individually or combine them.
Run this command to deploy the blueprint to your environment.
Runs your full outbound motion as one connected workflow using Claude Code and the Extruct API. It handles prospect research, company discovery, data enrichment, email generation, and campaign sending, with each skill sharing context so insights flow between stages instead of getting lost in disconnected tools.
Once installed via the skills CLI or Claude Code plugin manager, you describe your campaign goal in Claude. The skills work in parallel domains: build your context file and campaign hypotheses, discover target companies, enrich them with custom data, design and generate personalized emails, and manage sending and responses. Each skill reads from shared context and writes outputs that other skills consume. Output: a complete campaign from research through responses, all tracked in claude-code-gtm/ directories.
Foundation layer. Build and maintain your company context file and generate testable pain hypotheses.
Prospecting layer. Find companies matching your ICP, explore market verticals, and research competitive landscapes.
Data layer. Design enrichment columns, add custom research data, create tables, tier companies, and deep-dive accounts.
Intelligence layer. Locate decision makers and verified email addresses.
Content layer. Author email sequences, generate copies, and draft individual messages.
Execution layer. Validate addresses, quality-check emails, send campaigns, and manage responses.
Feedback layer. Track market signals, capture engaged audiences, and reference the Extruct API.
Every campaign artifact (context, lists, enrichment, emails, sends) stays in one project directory with shared context, so insights from enrichment inform email copy, and reply data feeds back into your hypotheses.
Requires Node 16+ for the skills CLI (npx skills). Skills are stored in ~/.claude/skills/ and work across all Claude Code projects. Set EXTRUCT_API_TOKEN in your terminal or provide it in chat. Email rendering is deterministic: it substitutes placeholders and applies HTML escaping as the final step, so template fixes apply across all rows.
Requires Extruct API token (free plan with 25 monthly credits; paid plans start at $49/mo for 500 credits). Skills that don't call the Extruct API (context-building, hypothesis-building, email-prompt-building, email-generation, atomic-message) work without it. Email sending and verification require integration with external providers (Instantly, contact enrichment services, etc.), and credentials are requested in chat when needed.
Context-Driven
All skills read from a single company context file (ICP, voice, win cases, proof points, active hypotheses) so every email and enrichment decision stays consistent.
API-Integrated Research
Search, lookalike discovery, and deep research via Extruct; enrich companies with custom AI-powered data points; find decision makers with LinkedIn profiles.
Flexible Campaign Paths
Start from your website, a win case, or a list of deals. Skills work independently, pick the workflow that fits your motion.
Templated Email Generation
Design email sequences once, then render them deterministically against any contact list without per-row LLM calls.
Response & Feedback Loop
Simulate prospect responses for quality review, manage replies in your inbox, and import campaign results back into your context file to refine hypotheses.