- Home
- Skills
- Mobile Development
- Adding React Native Screens to an Existing Native App
More screenshots
Works with the AI Agents you already use
Adding React Native Screens to an Existing Native App
Add one React Native screen to an existing native app: inspect the host, pick the shipping route, write the host-to-guest contract, and verify the release build.
$5
Adding React Native Screens to an Existing Native App
Example session with this skill installed
Our native running app, Ridgeline Run Club, already ships on iOS and Android. We want one new screen built with React Native: a weekly training plan that takes a runner's goal and current mileage and returns a chosen plan. One team owns both the native and React code in one repo. How do we add it without breaking the host app?
- Read your context and instructions
- Compiled the adding react native
HOST INSPECTION — Ridgeline Run Club (invented app): navigation owned by a coordinator on iOS and a single activity on Android; Gradle build on Android, manually maintained Xcode project on iOS; no React Native runtime linked yet; must-not-change list names the app object, the run-tracking flow, and the monthly release train. The manual Xcode project rules out any generation step against the host.
ROUTE DECISION — Integrated: one team owns native and React code in one repo, and the team accepts the JavaScript toolchain in the native build. Recorded reasons: single ownership, and one build to maintain.
CONTRACT — Input: goal text and current weekly mileage, serializable only. Result event: planChosen, fired exactly once, carrying plan id and start date. Dismissal: guest requests, coordinator performs. One listener, removed on unmount. Forbidden: own navigation, background work.
PROOF AND VERIFICATION — Throwaway screen: open with goal "first 10K" and 15 miles, return a plan, dismiss, reopen with goal "half marathon" and 25 miles. First run showed the old plan lingering; cause was a listener left registered, fixed at the boundary, second run clean. Release verification on the oldest supported test phone, bundle server stopped: plan screen opens from the training tab, returns the plan once, dismisses to the training tab, reopens fresh, no startup regression. Gate passed: 8 of 8 checklist items.
Connects securely to your tools. The creator never sees your data.
About this skill
Your native app already works, and the new screen should not put that at risk. This skill starts with the host, not the new code: a host inspection sheet covering navigation ownership, build system, signing, minimum operating system versions, and anything already linked, confirmed by the host owner before any choice is made. It then picks a shipping route from team facts — an isolated prebuilt artifact the native team consumes like a library, or an integrated build where one team owns everything — with a decision table and a rule that ownership breaks ties. Version discipline follows: keep the versions the project already selected, record them from the lockfile, and check the operating system floor before building. The heart of the skill is a five-clause host-to-guest contract — serializable input, a result event fired exactly once, host-performed dismissal, listener cleanup, and forwarded lifecycle events — written before the feature, then proved with a deliberately boring throwaway screen. The work is only done when a release build of the real host, with the development bundle server stopped, passes the acceptance run on the oldest supported device. Edge cases cover a host that already links a different runtime version, a guest that needs a new native capability, and deep links landing on the guest screen.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Mobile Development.
- It guides the integration and verification; it cannot run your builds or inspect your project directly.
- The release verification step needs a real device or simulator and a release build on your side.
- If the host cannot be inspected or its owner cannot confirm the must-not-change list, the route decision stays provisional until those facts are known.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 3 days ago
Permissions
No special permissions declared or detected