We train software developers and engineering teams to build production-ready software with AI agents, agent harness systems, rules, guardrails, tools, and memory — from prompt to deployment.
Obvious Works GmbH · Zug, Switzerland
We help software development teams adopt agentic coding — the practice of building software collaboratively with autonomous AI agents that plan, implement, test, and iterate independently.
Our focus areas:
- Agent Harness Architecture — designing the control layer (rules, commands, dos/don'ts) that makes AI agents consistent, secure, and reproducible
- Agentic Development Processes — establishing AI-first delivery loops with blast radius estimation, plan-by-conversation, guardrails, and stabilization
- MCP Integration — connecting AI agents to tools, databases, browsers, and APIs via the Model Context Protocol
- Context Engineering — token efficiency, structured markdown documentation, subsystem notes, and memory systems
- Multi-Agent Orchestration — coordinating agent swarms with collective memory and hive-mind topologies
- Quality & Safeguarding — 3-step review systems, TDD with AI agents, pre-commit security pipelines, and responsible AI practices
- Matthias - https://www.linkedin.com/in/matthiasherbert/
- Michael - https://www.linkedin.com/in/michaelmey/
We publish frameworks, rule sets, and best practices for the agentic coding community:
| Repository | Description |
|---|---|
| vibe-coding-ai-rules ⭐ 59 | The Ultimate Agentic Vibe Coding Guide — AI-optimized rules, best practices, and config files for Windsurf, Claude Code, Cursor, Codex & more |
| agentic-coding-rulebook | Production-ready collection of configuration files, templates, and best practices for professional AI-assisted coding across all major AI IDEs |
| agentic-coding-meta-prompt | The Meta-Prompt for Agentic Coding — synthesized from 1.5 MB of top global system prompts, combining APEX XML structure with ARCHITECT design |
| Claude-AI-skills-collection-2026 | Curated collection of official and community-built Claude Skills |
We offer hands-on training for developers, tech leads, and engineering teams — on-site, remote, or hybrid, in German and English.
| Training | Duration | Focus |
|---|---|---|
| Agentic Coding Hackathon | 3 days | Claude Code, MCP, agent harness systems, TDD with AI, RAG pipelines, multi-agent orchestration |
| AI Developer Bootcamp | 12 weeks | End-to-end AI development lifecycle — from harness design to production deployment |
| GenAI for Software Development | 2 days | AI-powered development workflows, coding assistants, and integration patterns |
| Training | Duration | Focus |
|---|---|---|
| AI Masterclass | 1 day | Generative AI overview — text, audio, video, design, automation |
| Getting Started with Generative AI | 1 day | Productivity boost with AI tools and LLMs |
| Grok Mastery | 1 day | Deep dive into xAI's Grok model |
| Product Owner 2.0: AI Superpower | 2 days | AI-driven product ownership and backlog management |
| Training | Duration | Focus |
|---|---|---|
| AI in Requirements Engineering | 2 days | LLM-assisted requirements elicitation, analysis, and validation |
| IREB AI4RE | 1 day | IREB-certified module on AI in requirements engineering |
| IREB CPRE Foundation | 3 days | IREB® Certified Professional for Requirements Engineering |
| CARS® — Agile Requirements Specialist | 2 days | Certified agile requirements engineering with AI prioritization |
- On-site (ImpactHub Zurich) · Live Remote · Hybrid
- German and English
- Max. 10–15 participants per cohort
- In-house training available for teams of 6+
- Agentic Coding Tools 2026: The 7 Frameworks That Take Your Development to a New Level
- Agent Harness: Why 90% of AI Agents Fail — and How You Can Do Better
- Claude Code Experience Report: 16 Hours Instead of 6 Weeks
- Token Optimization 2026: Save Up to 80% LLM Costs
- Master Prompt for AI Coding: The Ultimate System Prompt
- 🐅 Tiger Prompt Generator — Advanced prompt builder for ChatGPT and other LLMs
- 🤖 AI Coding Assistants Comparison 2025
- 🧭 LLM Navigator — Which AI for What
- 📘 User Story Hacks Booklet
Obvious Works GmbH Dammstrasse 16, CH-6300 Zug, Switzerland
📧 info[at]obviousworks.ch · ☎ +41 76 292 03 68