agentleFS
XinvtechXinvtech/mobile-design-os/skills/mobile-design-to-code/SKILL.md

mobile-design-to-code

Skill3 starsChanged 3 months agoMITView on GitHub

Convert validated mobile product, UX, wireframe, design-system, UI, prototype, and platform contracts into an implementation-ready Flutter handoff by default, or another explicitly requested stack. Use when Codex needs route/state mapping, token and component bindings, accessibility semantics, platform variants, implementation tasks, runnable scaffolding, or code-readiness review.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Plan: create a framework-specific implementation map and task breakdown before coding.

    line 12
  2. Distinctive:

    Scaffold: create a narrow runnable implementation for the declared platform and core flow.

    line 13
  3. Very specific:

    Prefer the validated Design Package plus wireframe.json, design-system.json, prototype.json, and platform-guidelines.json when available.

    line 20
  4. Distinctive:

    Read references/handoff-rules.md and references/handoff-schema.md before creating structured output.

    line 33
  5. Very specific:

    Run python3 scripts/validate_code_handoff.py <path-to-code-handoff.json> and fix every error.

    line 42
  6. Very specific:

    When a Flutter scaffold is requested, run python3 scripts/scaffold_flutter_app.py --handoff <path-to-code-handoff.json> --output …

    line 44
  7. Distinctive:

    Do not produce a screenshot-only handoff, invent inaccessible semantics, or claim implementation readiness when state, platform, or test …

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Mobile Design to Code

Preserve design intent as executable implementation contracts. Map meaning, states, behavior, and verification—not just screenshots or CSS-like coordinates.

Select the handoff mode

  • Plan: create a framework-specific implementation map and task breakdown before coding.
  • Scaffold: create a narrow runnable implementation for the declared platform and core flow.
  • Review: compare implementation against the design package, prototype tests, and platform contract.

Declare the target framework, OS versions, build entry point, persistence/network boundaries, and whether code changes are authorized. If the user has not selected a stack, use Flutter and record frameworkSelectionSource: project-default; switch to SwiftUI, UIKit, Jetpack Compose, React Native, or another stack only when the user explicitly asks for it, recording frameworkSelectionSource: user-specified.

Establish the input contract

Prefer the validated Design Package plus wireframe.json, design-system.json, prototype.json, and platform-guidelines.json when available. Preserve upstream IDs and provenance. Report missing contracts before writing code.

Build the mapping

  1. Map each design token to a semantic code token and source file.
  2. Map each component to a code component, props, state machine, accessibility semantics, and platform variant.
  3. Map each Screen/State to a route/view, data dependencies, loading/error/recovery behavior, and acceptance signals.
  4. Map gestures, motion, haptics, permissions, system surfaces, keyboard, lifecycle, and deep links to platform APIs or explicit fallbacks.
  5. Define test cases for core flows, state continuity, large text, assistive technology, reduced motion, offline, duplicate submission, persistence, and destructive recovery.
  6. Separate generated scaffolding from product decisions that require an engineer or designer.

For the project-default Flutter path, use references/flutter-implementation.md and the scaffold helper at scripts/scaffold_flutter_app.py. The helper creates a small runnable Flutter app from a valid handoff; it does not silently turn an untested prototype into production code.

Read references/handoff-rules.md and references/handoff-schema.md before creating structured output.

Produce and validate

Create:

  • code-handoff.md using assets/code-handoff-template.md.
  • code-handoff.json using assets/code-handoff-template.json.

Run python3 scripts/validate_code_handoff.py <path-to-code-handoff.json> and fix every error. A valid handoff is not proof that the app builds or that runtime behavior passes; keep implementation and device checks explicit.

When a Flutter scaffold is requested, run python3 scripts/scaffold_flutter_app.py --handoff <path-to-code-handoff.json> --output <new-project-directory>, then run flutter analyze and flutter test in the generated project. Keep those results separate from the static handoff evidence.

Enforce the output order

  1. Stack (Flutter by default unless explicitly overridden), scope, source contracts, and assumptions.
  2. Token/component bindings.
  3. Route/screen/state mapping.
  4. Platform and system API mapping.
  5. Tests, build/run instructions, ownership, and open risks.

Do not produce a screenshot-only handoff, invent inaccessible semantics, or claim implementation readiness when state, platform, or test contracts are missing.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.