run-evals
Installation
SKILL.md
Skill: Run Evals
Run the OpenWork UI evaluation flows against a real Electron app — either on a Daytona cloud sandbox or a local instance.
When to use
- User says "run evals on Daytona" or "run this flow on Daytona"
- User wants to verify a UI change end-to-end
- User wants to test the onboarding, session, or settings flows
Prerequisites
daytonaCLI installed and logged in (daytona login)- Using the "Different AI" org (
daytona organization use "Different AI") - The
.devcontainer/files exist in the repo
Workflow
Step 1: Create sandbox (if not running)
Related skills
More from different-ai/openwork
opencode-primitives
Reference OpenCode docs when implementing skills, plugins, MCPs, or config-driven behavior.
653solidjs-patterns
SolidJS reactivity + UI state patterns for OpenWork
558opencode-bridge
Bridge between OpenWork UI and OpenCode runtime
542tauri-solidjs
Tauri 2.x + SolidJS stack for OpenWork native app
506opencode-mirror
Maintain the local OpenCode mirror for self-reference
482openwork-core
Core context and guardrails for OpenWork native app
471