Catalog
Everything we ship.
Courses, templates, and community access — all in one place.
Courses
Free templates
Claude Skills Pack — production-ready skills, expanding monthly
A starter pack of 5 production-ready Claude Skills you can drop into Claude Code or the Agent SDK today: PDF-to-structured-data, meeting-notes extractor, Slack thread summarizer, codebase onboarder, weekly review. Every skill ships with SKILL.md + helper scripts. Monthly drops add new skills — your purchase is a lifetime subscription.
MCP Server Starter Pack — TypeScript templates
Two MCP server templates (TypeScript) wired against the Anthropic MCP spec: a Postgres read-only server and a generic REST-API wrapper. One npm install, instant Claude Code or Cursor integration. More starters (Slack, GitHub, S3) ship in monthly updates.
Cursor + Claude Code Rules Collection
Curated `.cursorrules` and `CLAUDE.md` libraries across the most-used stacks: Next.js App Router, FastAPI + Pydantic, n8n custom nodes, data-pipelines (Polars/DuckDB), and React Native. Each rule file is battle-tested on real projects — paste, ship, iterate. Quarterly updates add stacks.
Prompt Library Pro — 100+ field-tested prompts
100+ prompts organized by use case: sales outreach, ops automation, content multiplication, code review, research synthesis, decision frameworks. Ships as structured JSON (drop into your app) and Notion-importable CSV. PDF in v1.1, monthly drops add new prompts.
n8n AI Workflow Pack — production flows, monthly drops
Five hand-built n8n workflows you can import into a self-hosted or cloud n8n in one click: lead enrichment, inbox triage, content multiplication, RAG ingest, sentiment routing. Every workflow has a README explaining the env vars + rate-limit budget. Monthly drops add new flows.
AI-OS Dotfiles — opinionated setup for the AI engineer
Opinionated dotfiles + install script for the AI engineer's daily-driver: Claude Code, Cursor, zsh, tmux, fzf, bat, ripgrep, uv, mise. Sane defaults, project-aware MCP wiring, statusline that shows tokens/cost, one-line install on macOS or Linux.
RAG Starter Kit — Pinecone + Voyage + TS
Production-leaning RAG template: Pinecone vector store, Voyage embeddings, TypeScript ingest pipeline, and a small eval harness so you can measure retrieval quality before shipping. Drop your docs in, point it at your Claude/OpenAI key, and you have a real RAG endpoint in 30 minutes.
AI Voice Agent Starter — Twilio + Deepgram + Claude
End-to-end real-time voice agent: Twilio inbound number, Deepgram streaming STT, Claude turn-handling, ElevenLabs TTS. Designed for outbound qualification calls and inbound IVR replacement. Includes a barge-in handler and a transcript pipeline.
Multi-Agent Orchestration Pack — LangGraph patterns
Three production-shaped multi-agent patterns in LangGraph: a planner+executor split, a researcher+writer pair with shared scratchpad, and a parallel-fanout-with-judge for high-stakes outputs. Each ships with tests and a cost-budget guardrail.
Discord AI Bot Starter — moderation + onboarding + FAQ
A discord.js bot with three things every community wants: AI-assisted moderation (rule-based + Claude tiebreaker on edge cases), onboarding flow with automatic role assignment, and an FAQ assistant that pulls from your own /faqs channel. Drops onto Render or Railway in 10 minutes.
Slack AI Assistant — slash commands + thread DMs
A Slack app with `/ask`, `/draft`, `/summarize`, plus an in-thread DM assistant that reads the thread context and drafts replies. Bolt for JS, Socket Mode so you don't have to run an HTTPS endpoint, and a usage-cap config to protect your billing.
AI Newsletter Engine — sources → digest → send
Aggregate your favorite sources (RSS, X lists, Substacks, custom scrapers), run them through a Claude-powered digest with editor voice, send via Resend or Postmark. Includes draft preview email + a one-tap publish flow. Five worked source types in v1.
AI Customer Support Triage Kit — Zendesk + Intercom
Webhook receivers + classifier + auto-tagger for Zendesk and Intercom. Routes incoming tickets by sentiment + topic, drafts a first reply for the human to approve, and surfaces a daily 'top 5 themes' report. Ships with two real classifier prompts you can A/B.
WhatsApp Business AI Starter — Meta WA Cloud API
Meta WhatsApp Cloud API webhook receiver + Claude-powered reply engine + per-conversation memory in Postgres. Includes a template-message helper for the first outbound (Meta requires templates), and a drop-in handover flow when the user types 'human'.
Stripe Subscription AI Starter — webhooks + churn copilot
TypeScript Stripe webhook receiver wired to a Claude-powered churn copilot: classifies cancel reasons, drafts a tailored win-back email, and pushes annotated events to your CRM. Plus an idempotency layer so retried webhooks don't double-fire.
AI Sales Outreach Automation — research → draft → send
End-to-end outbound stack: account research (LinkedIn + company website + signals), Claude-drafted first email tuned to the recipient, and Resend / Postmark sender with reply detection. Designed for 50-200 outbound emails per week without losing the human touch.
The Vibe Coding Handbook — practical AI-assisted dev
An 80-page PDF + EPUB on building real software with Cursor, Claude Code, and the Agent SDK without losing your taste. Covers: prompt structure for code, when to let the agent run unattended, taming hallucinations, what to commit, and a 30-step refactor walkthrough.
RAG in Production — the 100-page handbook
What changes when your RAG goes from demo to production: chunk strategies that actually move retrieval@k, hybrid retrieval, freshness vs cost, eval pipelines, prompt-injection defense, observability, and the day-2 problems nobody warns you about. PDF + EPUB.
AI Cost Optimization Handbook — cut your bill 10× without losing quality
60-page handbook walking through every lever that actually moves AI spend: prompt-cache strategy, model-tier routing, output-token discipline, batch vs realtime, embedding choice, and three real before-after case studies cutting bills 8-12×. PDF + EPUB.
Unstructured RAG Ingest
Upload messy PDFs and chunk them into a Pinecone-ready retrieval flow without setting up the full paid stack first.
Sentiment Triage Flow
Detect urgent or frustrated user messages and route them into a human escalation path before the conversation slips.
AI Cost Calculator (Spreadsheet)
Plug your model, monthly request volume, prompt+response token estimates, and cache-hit rate into a Numbers/Excel spreadsheet — get a monthly USD bill. Pre-loaded with current Anthropic, OpenAI, and Voyage prices, plus per-tier guardrails.
AI Stack Decision Tree (PDF)
A 1-page decision tree that tells you, given a use case, which model family + retrieval strategy + agent pattern to start with. Built from 50+ shipped projects. Print it and stick it next to your monitor.
Prompt Audit Checklist (PDF)
Before you commit a prompt to production, run it past 18 questions: anchoring, format spec, escape hatches, refusal-on-uncertainty, output validation, and prompt-injection defense. Each question maps to one fix you can make today.
MCP Servers Comparison Sheet (Free)
An updated head-to-head spreadsheet of every meaningful MCP server we've tested: Postgres, Slack, GitHub, S3, Linear, Notion, JIRA, Sentry. Per-server: install command, auth model, tool surface, rate-limit ceiling, observability hooks, license, last-tested date.