SkillsTech Studio
Design and intent, for every role.
Product, design, and engineering describe what software should do, agree on it, and prove the work, all over one shared model. Studio is where you write it down and agree; IntentLang checks what you wrote; OpenThunder verifies the work; SkillsTech Workspace records approvals. Everything below runs offline, is deterministic (the same input always gives the same answer), and is honest about that line.
Who it is for: product managers, designers, researchers, engineers from entry level to senior, architects, QA, security, and reviewers. Everyone works from the same shared model, in language that fits their role.
Author intent
Three builders let anyone describe the product in the product's own terms, over the same shared model the rest of the team reads. Every evidence item is honestly classified, and an observed fact must carry a real source; the builder never assumes one.
Product Mission Builder
The problem, the evidence behind it, outcomes and their metrics, scope, non-goals, the never rules, open questions, and who must approve. It saves the same mission the Product view reads.
Experience Contract Builder
Journeys, screens and their states, interactions, and content. Accessibility and responsive assertions are always recorded as claims, never as verified.
Journey Builder
A user journey as an ordered path of steps: the action, the screen it lands on, the touchpoint, and any friction. A friction note is a hypothesis by default, never silently an observed fact.
A claim is proven by OpenThunder evidence and human review, not by writing it down.
Resolve, diff, and approve
When roles disagree, or an intent changes, Studio makes the disagreement and the impact visible, and keeps the decision with humans.
Conflict-Resolution Workspace
When two roles ask for things that cannot both be true, see the smallest set that conflicts, the ways to resolve it, and who must agree. A human chooses; a conflict is resolved only when every required role approves.
Semantic Diff
Compare two versions of an intent: impact grouped by role, the risks a change introduces, and a preview of which approvals it would invalidate. A comparison of the intent itself, not a verdict on the implementation.
Approval Workflow
Approvals are bound to the version they approved; when the intent moves forward, prior approvals go stale and need re-approval. AI can never grant an approval, and no one approves their own change.
Decisions are captured in Studio pending SkillsTech Workspace, which remains the system of record; the invalidation preview stays a preview until Workspace records the real thing.
A design system you can keep
Style Suggestions ranks polished style packs for your product, offline and with the same result every time, with a plain-language reason for each: no AI, no network. Preview a pack without changing anything, see a semantic Style Diff before you apply, and edit colors, typography, spacing, shape, and density by hand. Locked brand tokens and the approved logo cannot change without permission, so a style can never quietly replace your brand. Export to CSS variables or DTCG tokens.
Accessibility is asserted by each pack and verified by OpenThunder's Style Lens, never claimed here. AI-generated variations are a later, opt-in step, never required.
Friendly to every role
A first-class experience for newer contributors and non-engineers, not the engineering view with features removed.
Guided Contribution
Fifteen steps from the customer problem to verified proof, always showing the current step, why it matters, and what remains. It never fakes progress: a step that needs a real signal, a test run, a verification, a review, an approval, cannot be completed without it.
Explain and Teach
Any artifact, explained deterministically from the intent graph at the depth you want, in friendly, standard, or technical wording that never changes the meaning. An AI-enhanced explanation is optional and clearly marked; the deterministic one always stands on its own.
Role-adapted review on the web
The review console adapts its wording to your role, the same artifact presented for a product manager or an engineer, and lets a reviewer approve, request changes, or reject without version-control knowledge. AI never approves; the requester cannot approve.
Governance and operations
Decision tables, a lifecycle designer, authority and waivers, AI agent contracts, and telemetry contracts: Studio is where you author and preview them. Agent contracts state what an AI agent may touch, spend, and see; telemetry contracts state what is measured and what service levels mean. Tool scopes, budgets, data boundaries, and approval checkpoints are enforced at runtime by SkillsTech Runtime; approvals live in SkillsTech Workspace; policies are verified by OpenThunder, not by the editor.
The editor verifies and enforces nothing here: it authors the contract, and it says so.