agentleFS
ryanschueryanschue/automated-product-engineer/skills/parallel-platform-development/SKILL.md

parallel-platform-development

Skill0 starsChanged 24 days agoMITView on GitHub

Build one product feature concurrently in Next.js, native Kotlin Android, and native Swift iOS from a versioned feature capsule with Kotlin Multiplatform native business logic and machine-checked TypeScript/Kotlin semantic parity. Use for cross-platform monorepos, parallel platform delivery, contract generation, parity CI, or platform convergence planning.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read references/feature-capsule-protocol.md.

    line 12
  2. Distinctive:

    Keep authoritative business policy on the server.

    line 20
  3. Distinctive:

    Use KMP for the Android/iOS product kernel and portable runtime.

    line 22
  4. Very specific:

    Keep Next.js composition in React, Android composition in Jetpack Compose, and iOS composition in SwiftUI or UIKit.

    line 23
  5. Distinctive:

    Keep lifecycle, navigation, secure storage, capabilities, accessibility, and distribution native.

    line 24
  6. Very specific:

    Read references/parallel-operating-model.md.

    line 29
  7. Distinctive:

    Run the same scenario vectors in Kotlin and TypeScript.

    line 38
  8. Distinctive:

    Test platform-only behavior on its real runtime.

    line 41
  9. Very specific:

    Read references/adoption-and-debugging.md.

    line 48
  10. Distinctive:

    When parity fails, compare capsule digests first, then normalized inputs, policy outputs, serialization, numeric and time semantics, …

    line 50
  11. Distinctive:

    Update the capsule, generated projections, platform evidence, local documentation, and retrospective together.

    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

Parallel platform development

Ship one product decision through three native experiences without placing one platform on the critical path of another.

Start with a feature capsule

Read references/feature-capsule-protocol.md. Write the product language, actor and outcome, domain entities, states, intents, transitions, guards, authorization decisions, errors, telemetry, flags, tokens, component behavior, API schemas, and scenario vectors before platform implementation.

Assign the capsule a content digest. Every generated projection, platform branch, artifact, and evidence receipt must name that digest. A platform working against another digest is a different product version, even if the source compiles.

Preserve the ownership model

Read references/architecture-and-boundaries.md.

  • Keep authoritative business policy on the server.
  • Compile deterministic client-safe policy into Kotlin and TypeScript for validation, optimism, and explanation.
  • Use KMP for the Android/iOS product kernel and portable runtime.
  • Keep Next.js composition in React, Android composition in Jetpack Compose, and iOS composition in SwiftUI or UIKit.
  • Keep lifecycle, navigation, secure storage, capabilities, accessibility, and distribution native.
  • Share design meaning through tokens and component contracts, not rendered widgets by default.

Run independent lanes

Read references/parallel-operating-model.md. Create contract, server, shared-kernel, Next.js, Android, iOS, design, quality, and documentation lanes from the same capsule commit. Give each lane disjoint ownership and explicit inputs, outputs, proof, and integration boundary. Do not stack platform branches on each other.

Each platform publishes a projection receipt containing the capsule digest, artifact digest, supported capabilities, scenario results, focused tests, visual evidence, and known deviations. Integrate only receipts for the same capsule digest.

Prove convergence

Read references/parity-ci-and-release.md.

  1. Validate and regenerate contracts deterministically.
  2. Run the same scenario vectors in Kotlin and TypeScript.
  3. Compare canonical semantic receipts byte-for-meaning after parsing.
  4. Compile the KMP host, Android application, iOS framework plus Swift facade, and Next.js production application independently.
  5. Test platform-only behavior on its real runtime.
  6. Join the lanes at one convergence gate after all independent evidence exists.

Semantic parity does not replace native UX testing. It proves that platforms agree about product meaning; platform tests prove that each experience expresses it correctly.

Adopt incrementally

Read references/adoption-and-debugging.md. Begin with one high-value, concurrency-sensitive feature. Characterize existing behavior, create the capsule and scenario oracle, wrap current implementations, compare receipts, then move native shared logic into KMP. Preserve explicit deviations until they are product decisions.

When parity fails, compare capsule digests first, then normalized inputs, policy outputs, serialization, numeric and time semantics, optimistic behavior, and platform-only adapters. Fix the owner of the disagreement and add the failing case to the shared scenario vectors.

Finish

Update the capsule, generated projections, platform evidence, local documentation, and retrospective together. Report which decisions are shared, which remain native, what parity proves, remaining capability differences, and the next safe integration order.

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.