WorkflowOS

The Production AI Institute's reference implementation of the Production Safety Framework. Design workflows, score PSF controls, simulate runs, and export deployment evidence — free on the web or self-hosted from GitHub.

Open hosted Studio →Star on GitHub Why we open-sourced it →

Canvas

Six typed node kinds with contracts, validation notes, and auto-layout

PSF scoring

Eight-domain analysis on the graph you are actually shipping

Simulate

Deterministic run simulation and drift detection

AI assist

Wizard, heal, automate, executive brief — BYOK OpenAI

Export

YAML packages, proposals, diagrams, .wfos bundles

Evidence

Portfolio-ready exports for CPAP and client delivery

Reference implementation

Model the PSF in WorkflowOS

The standard is text. WorkflowOS is the working artifact — design workflows, run PSF analysis, simulate runs, and export evidence. Free on the web, or fork the MIT-licensed source and self-host for clients.

Open hosted Studio →View on GitHub

Self-host in four commands

The public repository contains Studio only — not the institute website, certifications, or billing stack. MSPs and integrators can deploy it on client infrastructure under the MIT licence.

git clone https://github.com/troygamble/workflowOS.git
cd workflowOS
npm install && cp .env.example .env.local
npm run dev

AI routes require Clerk sign-in and accept bring-your-own-key OpenAI via the Studio settings menu. See the README on GitHub for environment variables and security boundaries.