OpenLogos defines the standard.
RunLogos makes it land.
AI writes code faster than ever, and projects turn into black boxes. The LOGOS ecosystem pairs an open-source methodology that makes AI specify before it builds with a desktop workbench that lets you see, run and refine everything it produces.
The problem
The three costs of vibe coding
The faster AI writes, the faster these pile up
Ambiguity explodes
One vague request, ten interpretations. Whatever isn't written down becomes rework.
Black boxes accumulate
AI wrote 50 endpoints. Can you say why each exists — and whether it was ever verified?
Endless rework
Explaining to AI where to change three details takes longer than doing it yourself.
The LOGOS answer: specify first, then let AI build — every step traceable and verifiable.
Two products
One standard, one workbench
OpenLogos works standalone with any AI coding tool; RunLogos makes it effortless
Open-source methodology · CLI
OpenLogos
Apache 2.0An open-source software engineering methodology for the AI era.
- WHY → WHAT → HOW — each layer feeds the next
- Scenario-driven, test-first — tests define "done", three-layer traceability
- Documents as context — reviewable specs instead of black-box reasoning
- Skills + flow engine — works with Claude Code, Codex, Cursor, OpenCode
npm install -g @miniidealab/openlogos Desktop workbench
RunLogos
Desktop appThe best-practice tool for OpenLogos: visualize, orchestrate and debug everything AI generates.
- Visual documents — view & edit Markdown, prototypes, OpenAPI, DB schemas
- API orchestration — run AI-written test scenarios and debug step by step
- Comment-to-prompt — turn review feedback into precise AI instructions
- Works with your tools — Cursor, Claude Code, OpenCode, Codex
Inside RunLogos
See it, run it, refine it
Turn OpenLogos specs into a workbench you can act on
Every AI-written spec, at a glance
Requirements, prototypes, OpenAPI and schemas in one desktop app — browse, compare and edit, always knowing the real state of your project.
How they fit
Methodology → tool → AI → methodology
A development loop that keeps converging
OpenLogos
Specify first
Produce requirements, design, scenarios and test cases, WHY → WHAT → HOW.
AI coding tools
AI builds to spec
Skills guide your AI tool to write code and tests; tests define done.
RunLogos
See and run
Review specs visually, orchestrate API tests, verify real flows.
RunLogos → OpenLogos
Feed back into specs
Comments become prompts; changes flow back into specs through traceable proposals.
Pick your starting point
Start with the open methodology, or grab the desktop workbench. Docs, releases and downloads live on each product site.