Junior → Mid RN Leveling

A cookbook skills matrix for mobile-specific competencies on Expo SDK 57 teams - define what junior looks like, what mid owns, and how to close gaps with evidence.

Recipe

Quick-reference leveling card - copy into eng ladder or performance rubric.

Junior RN (L3) - "ships with support"
  ✓ Expo Router screens + Link navigation
  ✓ RNTL tests for happy path
  ✓ PR checks green; fixes own lint/tsc failures
  ✓ Runs app on sim + one physical device
  ✓ Asks for review before touching native config
 
Mid RN (L4) - "owns vertical slices"
  ✓ Feature folder end-to-end: route → API → test → preview build
  ✓ TanStack Query patterns; no duplicate fetch in global store
  ✓ Debugs Metro, Hermes redboxes, and flaky CI
  ✓ Triggers preview EAS build; reads eas.json profiles
  ✓ Authors 1–3 Maestro smoke steps or equivalent E2E
  ✓ Participates in on-call / release rotation with checklist

When to reach for this:

Calibrating promotion discussions beyond vague "strong engineer."
Building growth plans for web-to-mobile transfers.
Staffing tickets - juniors on bounded UI; mids on cross-layer features.

Working Example

Competency matrix

Rate A (autonomous), G (guided), N (not yet) per quarter. Mid bar = A on all mid rows, G+ on most junior rows.

Competency	Junior (L3)	Mid (L4)	Evidence
Expo Router	G - adds screens in existing groups	A - route groups, modals, typed params	Merged PRs touching `app/`
React / RN UI	G - lists, forms, loading states	A - perf-aware lists, keyboard, a11y roles	Reviewer notes; RNTL queries by role
TypeScript	G - props types, fixes tsc errors	A - narrow unions at route boundaries	Zero tsc suppressions in owned files
Testing	G - one component test per PR	A - hook tests + MSW integration	CI coverage on owned feature folder
Networking	G - calls existing API hooks	A - designs Query keys, error UX, retry policy	TanStack Query on Mobile patterns
Navigation edge cases	N - back stack surprises	G - deep links, auth redirects	Manual test matrix documented in PR
Native / EAS	N - uses provided dev client	G - triggers preview build, reads logs	EAS dashboard screenshot in PR
Permissions	N - copies existing pattern	G - camera/location flows with denial path	Screenshots of OS dialogs in QA notes
Offline / NetInfo	N - aware	G - empty/offline UI states	Unit test with mocked NetInfo
Performance	N - avoids obvious rerenders	G - FlashList, memo boundaries	Profiler before/after on hot list
CI/CD literacy	G - fixes PR check failures	A - knows preview vs release lanes	Mobile CI/CD Basics quiz pass
Release hygiene	N	G - runtimeVersion awareness, no secrets in bundle	Checklist sign-off on RC
Debugging	G - Metro cache, basic redbox	A - symbolicated traces, device logs	Postmortem participation
Communication	G - daily standup, asks early	A - estimates slices, flags native risk in planning	Ticket breakdown quality

Junior → mid gap plan (12 weeks)

One competency focus per sprint - not "learn everything."

## <Engineer> - Q3 leveling plan
 
| Sprint | Focus | Success signal |
| --- | --- | --- |
| 1 | Router + thin routes | 2 screens via feature re-export |
| 2 | RNTL by role/label | 80% new UI has test |
| 3 | TanStack Query reads | Remove one Zustand fetch duplication |
| 4 | Error + loading UX | Standard ErrorView on owned feature |
| 5 | Preview EAS build | Solo `eas build --profile preview` |
| 6 | Physical device QA | Device matrix row in PR template |
| 7 | Deep link or auth redirect | One `Redirect` / guard flow owned |
| 8 | Maestro smoke slice | 3-step YAML flow in CI optional job |
| 9 | List performance | Migrate one ScrollView → FlashList |
| 10 | Offline empty state | NetInfo mock test + UI |
| 11 | Release shadow | Buddy on RC checklist |
| 12 | Mid packet review | Matrix all A/G on mid rows |

Ticket sizing guide

Level	Appropriate tickets	Avoid until mid
Junior	New row UI, copy, icon, isolated hook, test fixes	SDK bump, new config plugin, store submit
Mid	New feature slice, API integration, navigation restructure	Solo SDK major upgrade without staff engineer
Senior+	Cross-app modules, native bridges, CI architecture	-

Junior ticket example:
  "Add order status badge to OrderRow - uses existing useOrder hook"
 
Mid ticket example:
  "Add order cancellation flow - new route, mutation, error states, preview build"

Web transfer acceleration map

Web skill	Maps to mobile	Gap to close
React hooks	Same mental model	RN primitives (`View`, `Pressable`)
CSS layout	Flexbox similar	No cascade - StyleSheet / NativeWind
React Router	Expo Router file routes	Imperative `router.push`, mobile back stack
Browser DevTools	React DevTools	Flipper / React Native DevTools, sim logs
`fetch` + SWR	TanStack Query	Offline, app backgrounding, token refresh
Jest + RTL	RNTL	`jest-expo`, native module mocks
Deploy to Vercel	EAS Build + stores	Code signing, phased rollout

Promotion packet (evidence list)

Mid promotion should cite artifacts, not adjectives:

## Mid promotion evidence - <name>
 
### Delivery
- 3 feature slices owned E2E (link PRs)
- 1 preview-build-driven QA cycle led
 
### Quality
- RNTL tests on owned features - <count> tests added
- Zero P1 bugs traced to missing loading/error state in owned code
 
### Mobile craft
- Maestro smoke: <link or flow file>
- Device matrix completed on RC <version>
 
### Collaboration
- Onboarded 1 hire on codebase tour
- Release checklist buddy on <date>
 
### Growth edges (honest)
- Still G on: push notification cold start
- Plan: pair with <senior> in Q4

Deep Dive

Signals that are not leveling

Signal	Why it fails
"Knows React for 5 years"	Web tenure ≠ native release literacy
"Merged 100 PRs"	Count without scope - typo fixes ≠ feature ownership
"Works late"	Hours ≠ autonomy on EAS or device matrix
"Strong opinions on Redux"	Architecture debates without shipping evidence

Signals that are leveling

Autonomous preview build when native-adjacent UI changes
Query cache designed so sibling screens do not double-fetch
Reviewer trust on owned feature folder - fewer nit comments over time
Incident participation with actionable notes ("missing offline gate")

Staff calibration questions

Use in leveling meetings:

Can they ship a feature without daily routing help?
Do their PRs include device evidence when gestures or safe areas change?
Can they explain why OTA cannot fix their last change?
Did they catch a native risk in planning before sprint commit?

Recommended learning by gap

Gap	Resource in this docs site
Router	Expo Router Basics
EAS	EAS Build Basics
Testing	Mobile Testing Basics
CI	Mobile CI/CD Basics
Structure	Folder Structure for Features
Performance	Performance Basics

FAQs

Junior with strong web background - start at mid?

UI velocity may be mid-level; native/EAS/permissions likely junior. Split matrix rows - do not skip device and release literacy.

How many mid engineers per squad?

Common ratio: 2 mid : 1 senior : 1–2 junior for a product squad of 4–5. Adjust when release load is high.

Contractor on 6-month term?

Use matrix for scope boundaries, not promotion. Contractors can be "mid scope" without L4 title.

Expo Go only team - defer EAS?

No - mid bar still requires preview build literacy. Use org's shared dev client if personal EAS access is limited.

Onboarding Basics - day-one setup
First-Week Learning Path - week-one capstone
Codebase Tour Checklist - repo orientation
Project Setup Basics - SDK 57 conventions
Expo Router Basics - navigation competency
EAS Build Basics - build literacy
Team & Onboarding Best Practices - 25 summary practices

Stack versions: This page was written for React 19.2.3, React Native 0.86.0, and Expo SDK 57 (expo ~57.0.4).

Recipe

Quick-reference leveling card - copy into eng ladder or performance rubric.

Junior RN (L3) - "ships with support"
  ✓ Expo Router screens + Link navigation
  ✓ RNTL tests for happy path
  ✓ PR checks green; fixes own lint/tsc failures
  ✓ Runs app on sim + one physical device
  ✓ Asks for review before touching native config
 
Mid RN (L4) - "owns vertical slices"
  ✓ Feature folder end-to-end: route → API → test → preview build
  ✓ TanStack Query patterns; no duplicate fetch in global store
  ✓ Debugs Metro, Hermes redboxes, and flaky CI
  ✓ Triggers preview EAS build; reads eas.json profiles
  ✓ Authors 1–3 Maestro smoke steps or equivalent E2E
  ✓ Participates in on-call / release rotation with checklist

When to reach for this:

Calibrating promotion discussions beyond vague "strong engineer."
Building growth plans for web-to-mobile transfers.
Staffing tickets - juniors on bounded UI; mids on cross-layer features.

Working Example

Competency matrix

Rate A (autonomous), G (guided), N (not yet) per quarter. Mid bar = A on all mid rows, G+ on most junior rows.

Competency	Junior (L3)	Mid (L4)	Evidence
Expo Router	G - adds screens in existing groups	A - route groups, modals, typed params	Merged PRs touching `app/`
React / RN UI	G - lists, forms, loading states	A - perf-aware lists, keyboard, a11y roles	Reviewer notes; RNTL queries by role
TypeScript	G - props types, fixes tsc errors	A - narrow unions at route boundaries	Zero tsc suppressions in owned files
Testing	G - one component test per PR	A - hook tests + MSW integration	CI coverage on owned feature folder
Networking	G - calls existing API hooks	A - designs Query keys, error UX, retry policy	TanStack Query on Mobile patterns
Navigation edge cases	N - back stack surprises	G - deep links, auth redirects	Manual test matrix documented in PR
Native / EAS	N - uses provided dev client	G - triggers preview build, reads logs	EAS dashboard screenshot in PR
Permissions	N - copies existing pattern	G - camera/location flows with denial path	Screenshots of OS dialogs in QA notes
Offline / NetInfo	N - aware	G - empty/offline UI states	Unit test with mocked NetInfo
Performance	N - avoids obvious rerenders	G - FlashList, memo boundaries	Profiler before/after on hot list
CI/CD literacy	G - fixes PR check failures	A - knows preview vs release lanes	Mobile CI/CD Basics quiz pass
Release hygiene	N	G - runtimeVersion awareness, no secrets in bundle	Checklist sign-off on RC
Debugging	G - Metro cache, basic redbox	A - symbolicated traces, device logs	Postmortem participation
Communication	G - daily standup, asks early	A - estimates slices, flags native risk in planning	Ticket breakdown quality

Junior → mid gap plan (12 weeks)

One competency focus per sprint - not "learn everything."

## <Engineer> - Q3 leveling plan
 
| Sprint | Focus | Success signal |
| --- | --- | --- |
| 1 | Router + thin routes | 2 screens via feature re-export |
| 2 | RNTL by role/label | 80% new UI has test |
| 3 | TanStack Query reads | Remove one Zustand fetch duplication |
| 4 | Error + loading UX | Standard ErrorView on owned feature |
| 5 | Preview EAS build | Solo `eas build --profile preview` |
| 6 | Physical device QA | Device matrix row in PR template |
| 7 | Deep link or auth redirect | One `Redirect` / guard flow owned |
| 8 | Maestro smoke slice | 3-step YAML flow in CI optional job |
| 9 | List performance | Migrate one ScrollView → FlashList |
| 10 | Offline empty state | NetInfo mock test + UI |
| 11 | Release shadow | Buddy on RC checklist |
| 12 | Mid packet review | Matrix all A/G on mid rows |

Ticket sizing guide

Level	Appropriate tickets	Avoid until mid
Junior	New row UI, copy, icon, isolated hook, test fixes	SDK bump, new config plugin, store submit
Mid	New feature slice, API integration, navigation restructure	Solo SDK major upgrade without staff engineer
Senior+	Cross-app modules, native bridges, CI architecture	-

Junior ticket example:
  "Add order status badge to OrderRow - uses existing useOrder hook"
 
Mid ticket example:
  "Add order cancellation flow - new route, mutation, error states, preview build"

Web transfer acceleration map

Web skill	Maps to mobile	Gap to close
React hooks	Same mental model	RN primitives (`View`, `Pressable`)
CSS layout	Flexbox similar	No cascade - StyleSheet / NativeWind
React Router	Expo Router file routes	Imperative `router.push`, mobile back stack
Browser DevTools	React DevTools	Flipper / React Native DevTools, sim logs
`fetch` + SWR	TanStack Query	Offline, app backgrounding, token refresh
Jest + RTL	RNTL	`jest-expo`, native module mocks
Deploy to Vercel	EAS Build + stores	Code signing, phased rollout

Promotion packet (evidence list)

Mid promotion should cite artifacts, not adjectives:

## Mid promotion evidence - <name>
 
### Delivery
- 3 feature slices owned E2E (link PRs)
- 1 preview-build-driven QA cycle led
 
### Quality
- RNTL tests on owned features - <count> tests added
- Zero P1 bugs traced to missing loading/error state in owned code
 
### Mobile craft
- Maestro smoke: <link or flow file>
- Device matrix completed on RC <version>
 
### Collaboration
- Onboarded 1 hire on codebase tour
- Release checklist buddy on <date>
 
### Growth edges (honest)
- Still G on: push notification cold start
- Plan: pair with <senior> in Q4

Deep Dive

Signals that are not leveling

Signal	Why it fails
"Knows React for 5 years"	Web tenure ≠ native release literacy
"Merged 100 PRs"	Count without scope - typo fixes ≠ feature ownership
"Works late"	Hours ≠ autonomy on EAS or device matrix
"Strong opinions on Redux"	Architecture debates without shipping evidence

Signals that are leveling

Autonomous preview build when native-adjacent UI changes
Query cache designed so sibling screens do not double-fetch
Reviewer trust on owned feature folder - fewer nit comments over time
Incident participation with actionable notes ("missing offline gate")

Staff calibration questions

Use in leveling meetings:

Can they ship a feature without daily routing help?
Do their PRs include device evidence when gestures or safe areas change?
Can they explain why OTA cannot fix their last change?
Did they catch a native risk in planning before sprint commit?

Recommended learning by gap

Gap	Resource in this docs site
Router	Expo Router Basics
EAS	EAS Build Basics
Testing	Mobile Testing Basics
CI	Mobile CI/CD Basics
Structure	Folder Structure for Features
Performance	Performance Basics

FAQs

Junior with strong web background - start at mid?

UI velocity may be mid-level; native/EAS/permissions likely junior. Split matrix rows - do not skip device and release literacy.

How many mid engineers per squad?

Common ratio: 2 mid : 1 senior : 1–2 junior for a product squad of 4–5. Adjust when release load is high.

Contractor on 6-month term?

Use matrix for scope boundaries, not promotion. Contractors can be "mid scope" without L4 title.

Expo Go only team - defer EAS?

No - mid bar still requires preview build literacy. Use org's shared dev client if personal EAS access is limited.

Onboarding Basics - day-one setup
First-Week Learning Path - week-one capstone
Codebase Tour Checklist - repo orientation
Project Setup Basics - SDK 57 conventions
Expo Router Basics - navigation competency
EAS Build Basics - build literacy
Team & Onboarding Best Practices - 25 summary practices

Stack versions: This page was written for React 19.2.3, React Native 0.86.0, and Expo SDK 57 (expo ~57.0.4).

Junior → Mid RN Leveling

Search Documentation

Junior → Mid RN Leveling