AI Developer Cards box in a neon-lit developer workspace
session://ai-developer-cards

AI Developer Cards — build_better
software with_AI

For developers, team leads and trainers.
9 quests and 45+ methods for building, modernizing and operating software with AI. Learn to give coding agents clear tasks and check their results.

9
quests
45+
methods
100+
workshops
1
card deck

AI Developer Cards is a card deck and companion app by Ingo Eichhorst and Ralf D. Mueller. Developers use the methods on their own tasks. Team leads use them to establish shared practices. Trainers use them for workshops. The deck covers 45+ methods in nine quests, from prototypes to production. It draws on experience from 100+ developer workshops.

The deck.

click to flip · hover to inspect

Cheat codes that survived 100+ workshops.
Zero fluff.

AI Developer Cards box in a developer workspace
01 // tactile_box_set

Large-format physical cards

Premium card stock, oversized format. 9 quest cards and 45+ method cards covering every scenario from first prototype to production hardening.

AI Developer Cards companion app showing a card, the card browser, and the story
02 // companion_app

Story and digital cards

Read PRESS START in the companion app. Follow developer Alex and his coding agent Sprite. The story shows you when each method helps. Open the relevant cards, learn more with articles and references, try learning exercises and copy prompts for your work.

Two Model Selection cards with rounded borders showing an update from v1.1 to v1.2
03 // always_current

Evergreen knowledge base

Constant app updates and discounted booster packs keep your toolbox aligned with the latest LLM capabilities, toolchains, and best practices.

The quest map.

Each quest is a focused mission — a curated sequence of methods from problem to production.

Q0 5 methods

ai_prototyping

Build a prototype with AI. Check what the tools can do and where you need to decide.

Q1 5 methods

spec_driven_development

Develop an application from a clear specification. Keep the code tested and easy to change.

Q2 5 methods

agentic_coding

Let coding agents plan tasks, change code, and check their results.

Q3 5 methods

legacy_modernization

Understand, plan, and change existing systems step by step with AI.

Q4 5 methods

multi_agent_coding

Coordinate several coding agents with clear tasks, separate workspaces, and shared checks.

Q5 5 methods

harness_engineering

Shape the codebase, tools, and feedback so agents work reliably within clear limits.

Q6 5 methods

software_factories

Build a measurable software process with bounded agent tasks and automated and human checks.

Q7 5 methods

reinforcement_improvement

Use previous agent runs and measured feedback to improve rules, skills, and checks.

Q8 5 methods

ai_software_hardening

Limit dangerous agent actions. Prepare controls, audits, incident response, and safe changes.

Be first in line.

Early-bird pricing, beta app access, and an exclusive preview PDF — no spam, unsubscribe any time.

Early-bird pricing on the physical deck Beta access to the companion app Exclusive preview PDF No spam — unsubscribe any time
$ join_waitlist --early-access

By joining, you agree to our Privacy Policy. Data collected via Google Forms.

Ingo Eichhorst
ingo_eichhorst // ai_architect

15+ years.
One playbook.

AI architect, CTO, and engineering trainer at a leading hyperscaler. Beta tester for GitHub Copilot and GPT-3 in 2021 — it clicked immediately. After training thousands of developers across Europe and collecting countless feedback loops, I distilled what actually moves the needle. These cards are the playbook I wish I had from day one.

→ 100+ AI developer workshops delivered across Europe
→ Open-source contributor & university lecturer
→ Speaker at betterCode(), AI Camp & IT-Tage

Ralf D. Müller
ralf_d_mueller // software_architect

25+ years.
Docs as Code.

Software architect, arc42 & docToolchain creator, and GenAI pioneer. Co-host of Software-Architektur.TV and CPSA-A certified trainer at socreatory. After 25+ years in software development, Ralf bridges the gap between architecture documentation and modern AI-assisted workflows — teaching teams to speak the same language as their LLMs through semantic anchors.

→ Creator of docToolchain & arc42 template maintainer
→ Co-host Software-Architektur.TV & GenAI trainer at socreatory
→ Published in iX Magazine on AI in software architecture

Frequently Asked Questions

I already use AI coding tools every day — what am I actually missing?

Most developers have mastered the basics: autocomplete, Q&A, simple refactoring. What breaks down is knowing which method to reach for at each stage — when to vibe-prototype vs. harden production code, how to engineer context for consistent results, how to chain methods across a full workflow. AI Developer Cards map 45+ methods across 9 progression stages, distilled from 100+ workshops, so you always know the right move — not just the right prompt.

What did running 100+ developer workshops reveal?

A consistent pattern: developers kept reinventing the same mental models from scratch. After 100+ AI coding workshops across Europe, Ingo Eichhorst and Ralf D. Mueller noticed that the gap wasn't in AI capability — it was in structured methodology. Most teams got stuck cycling between basic autocomplete and ad-hoc prompting with no clear progression. That insight became the blueprint for the 9 quests and 45+ methods in the deck.

What AI coding methods do the cards cover?

The deck covers 45+ methods across 9 quests: AI Prototyping, Spec Driven Development, Agentic Coding, Legacy Modernization, Multi-Agent Coding, Harness Engineering, Software Factories, Reinforcement Improvement, and AI Software Hardening. Each method card includes specific prompts, patterns, and mental models for applying AI tools in real development scenarios.

Who are AI Developer Cards for?

The cards are for developers, team leads and trainers. Use the methods on your own coding tasks, establish shared practices in your team or lead a workshop. The nine quests guide you from your first prototype to production. Learn with the PRESS START story in the companion app or look up a method for your current problem.

What is included in the card deck?

The physical deck contains nine quest cards and 45+ method cards. The companion app includes the complete PRESS START story, the latest card versions, articles, references, learning exercises and prompts you can copy. Start with the story or look up a method directly.

When will AI Developer Cards be available?

AI Developer Cards are currently in pre-launch. Waitlist members receive 48-hour early access before any public announcement. Join above to get early-bird pricing on the physical deck, beta access to the companion app, and an exclusive preview PDF as soon as the deck ships.