How are these different from free articles on the site?

Free articles cover individual topics. Products bundle the full system in one organized place plus material not published elsewhere — failure modes, cost case studies, complete code.

Digital Products

Q: What format do I get?

Every product is delivered as a printable PDF. Power Prompts also includes JSONL for direct Claude Code import; Cost Optimization includes a 6-sheet Excel calculator. Instant download.

Practical ebooks, prompt packs, and playbooks for developers building with Claude. Instant download via Gumroad.

✓ 30-day money-back guarantee✓ Instant PDF download✓ Free updates for 12 months✓ 5 verified case studies

Not sure yet? Try free first: 30 prompts · cost calculator · case studies · all 16 free resources

Which product is right for me?

If you're...	Start with	Price
Spending too much on Claude API ($1K+/mo)	P5: Cost Optimization Masterclass	$59
Using Claude Code daily, want better prompts	P1: Power Prompts 300	$29
Building production agents with the SDK	P2: Agent SDK Cookbook	$49
Want AI engines (Claude/ChatGPT) to cite your content	P3: AEO Playbook	$39
Solo dev figuring out the AI tool stack	P4: Solo AI Builder Stack	$19

Claude API Cost Optimization Masterclass

Best value

Cut your Claude API bill by 60–90% without sacrificing quality.

12 production deployments analyzed. The concrete order-of-operations to pull the biggest levers first: prompt caching, model tiering, Batch API, token compression.

✓ 120-page PDF + 6-sheet Excel cost calculator
✓ Module 1: Prompt Caching — up to 90% savings
✓ Module 2: Model Tiering — Haiku vs Sonnet vs Opus
✓ Module 3: Batch API — 50% discount for async workloads
✓ Real results: $2,100 → $187/month on customer support agent
✓ 30-day money-back guarantee

Claude Code Power Prompts 300

300 battle-tested prompts. Copy, paste, ship.

Stop reinventing prompts every morning. Every prompt tested in production Claude Code sessions, organized by use case, with exact CLAUDE.md context and expected output.

✓ 300 copy-paste-ready prompts
✓ Organized: debugging, refactoring, testing, architecture, security
✓ 40 slash command templates
✓ Token-optimized variants (save up to 70% on API costs)
✓ JSONL file for direct Claude Code import
✓ 12 months of free updates
✓ 30-day money-back guarantee

Claude Agent SDK Cookbook: 40 Production Patterns

New

The pattern library for agents that run unsupervised in production.

Not a getting-started guide. 40 battle-tested patterns for the Claude Agent SDK: retry logic, tool error recovery, parallel sub-agents, cost guardrails, deterministic testing. Python and TypeScript code throughout.

✓ 167-page PDF with complete, runnable code
✓ 40 production patterns with failure modes + recovery logic
✓ Multi-agent orchestration — supervisor patterns, agent handoffs
✓ Cost controls — model routing, caching, batch processing
✓ Real-world recipes: web scraping, code review, data pipeline agents
✓ 30-day money-back guarantee

AEO Playbook: Rank in AI Answers (Claude, ChatGPT, Gemini)

New

Stop optimizing for Google. Start optimizing for the AI answer layer.

ChatGPT, Claude, and Gemini are the new search. This playbook is the complete system for getting cited, quoted, and recommended by AI engines — audit framework, content templates, and visibility tracking.

✓ 134-page PDF with step-by-step implementation guides
✓ AEO audit framework — diagnose why AI ignores your content
✓ Claude-specific optimization + ChatGPT & Gemini signals
✓ 12 content templates (copy-paste into your CMS)
✓ AEO audit checklist (Google Sheets compatible)
✓ 30-day money-back guarantee

Solo AI Builder Stack

The operating system for building one AI product that makes money.

Not a productivity system. A 20-page PDF + Notion workspace template built from a real 12-month, 10M KRW AI monetization sprint. Budget model, marketing framework, distribution system.

✓ 20-page PDF companion guide
✓ Notion workspace template (duplicate link included)
✓ Solo Builder's Budget Model — the one number that runs everything
✓ 3-layer marketing system that works without an audience
✓ Korean solo builder context: banking, taxes, platforms
✓ Revenue tracking dashboard template
✓ 30-day money-back guarantee

Frequently asked

How does the 30-day money-back guarantee work?

If a product doesn't deliver what was promised, email firstshout@gmail.com within 30 days of purchase. No questions asked. Gumroad processes the refund directly back to your card, usually within 5-7 business days.

What format do I get?

Every product is delivered as a printable PDF (Power Prompts also includes JSONL for direct Claude Code import; Cost Optimization includes a 6-sheet Excel calculator). Instant download after purchase — no account required.

How are these different from free articles on this site?

The free articles cover individual topics. The products bundle the full system — every prompt, every pattern, every checklist in one organized place — plus material that isn't published elsewhere (failure modes, cost case studies, complete code).

Are updates included?

Yes. Every product gets free updates for 12 months. When Claude models change or new patterns emerge, you receive the new edition at no charge via the same Gumroad link.

Can I expense these as a developer?

Yes — Gumroad provides a receipt suitable for reimbursement. Many readers expense Power Prompts 300 ($29) and Cost Optimization ($59) as professional development.

Is there a bundle discount?

Not currently. If you want all five, the total is $195 — but start with one and refund if it doesn't deliver. Most readers find the first product pays for itself within a week.

Why pay when Anthropic's docs are free?

Anthropic's docs are reference material. Our products are opinionated systems: a tested order-of-operations to cut costs (P5), a 300-prompt curated library with tested variants (P1), and 15 production agent recipes with retry middleware and cost guardrails (P2). The free articles on this site cover individual topics. The products bundle the full system in organized form.

Will this save me money or just teach me?

5 case studies show real before/after monthly costs: $2,100 → $187 (91%), $4,800 → $640 (87%), $5,500 → $1,150 (79%). Most P5 buyers pay back the $59 within their first month of API spend if they're running >=$200/month on Claude.

What if I'm using Bedrock or Vertex AI?

Patterns work the same way (model routing, caching with cache_control, Batch API). Bedrock-specific notes are included in P5. Direct API typically remains 10-15% cheaper than Bedrock for the same model.

Are these tested in production?

Yes. Code samples are run against Anthropic's API before ship. Cost numbers in P5 come from real customer engagements (anonymized). See /benchmarks for the verified, dated data points across the site.

Will this become outdated when new Claude models release?

Models change, but the patterns don't (Haiku/Sonnet/Opus tier shape persists, caching mechanics persist, Batch API persists). Free updates for 12 months means you receive every revision when models or pricing change.

All products are independent publications. Questions? firstshout@gmail.com

Also tracking your Claude API costs? claudecosts.app — free dashboard for your Anthropic Admin API usage. Connect your key once, see daily spend by model. No credit card.