Stop Vibe Coding.
Start Building Right.
OpenLogos is an open-source software engineering methodology for the AI era. It turns AI coding tools from guessing machines into precision instruments — built on four core principles.
npx @miniidealab/openlogos init my-project Most AI coding today is "Vibe Coding"
Jump straight to code and hope for the best. The result? Ambiguity explosion, black-box accumulation, and unmaintainable projects that collapse at scale.
Ambiguity Explosion
Without clear requirements, AI guesses what you want. Every guess compounds uncertainty.
Black-Box Accumulation
AI generates code you don't understand. Over time, your codebase becomes opaque.
Endless Rework
No specs means no definition of "done." You iterate forever without converging.
Four pillars that make AI
actually reliable
OpenLogos is built on battle-tested software engineering theory, re-compiled for the AI era. Each principle reinforces the others.
WHY → WHAT → HOW
Three progressive phases. Each layer's output feeds the next. Skip one and ambiguity explodes.
Learn moreScenario-Driven + Test-First
One scenario ID traces from requirements to code. Tests define "done" before a line is written. Quality is a number, not a feeling.
Learn moreDocuments as Context
In Vibe Coding, AI's context is a black box. In OpenLogos, every decision is an open document — reviewable, reproducible, cumulative.
Learn moreEngineering Foundation
Not invented from scratch. Built on 40 years of proven theory — BDD, TDD, DDD, Stage-Gate — recompiled for AI execution.
Learn moreSay one sentence. AI does the rest.
OpenLogos encodes methodology as AI Skills — Markdown files that guide your AI tool through each development phase automatically.
AI reads AGENTS.md, runs openlogos init, scaffolds the entire project structure with configuration, Skills, and directory hierarchy.
AI detects Phase 1, loads the prd-writer Skill, interviews you about pain points and scenarios, generates a structured requirements doc with GIVEN/WHEN/THEN.
AI loads api-designer Skill, reads the sequence diagram, generates OpenAPI YAML spec with request/response schemas and error codes.
AI designs test cases first (test-first!), then generates implementation code with full context — prototype + scenarios + API + DB + tests.
Runs openlogos verify — three-layer traceability checks pass rate, design-time coverage, and acceptance criteria. Gate pass = scenario done.
Works with your AI tools
Skills are platform-agnostic Markdown files. Any AI tool that can read project files works with OpenLogos.
Up and running in 2 minutes
# Install the CLI
$ npm install -g @miniidealab/openlogos
# Initialize your project
$ openlogos init my-project
✓ Created logos/logos.config.json
✓ Created logos/logos-project.yaml
✓ Created AGENTS.md
✓ Project "my-project" initialized
# Open in your AI tool and start building
$ cd my-project && claudeReady to build software
the right way?
Join the community. Contribute skills. Shape the future of AI-assisted development.