Code Review

Code Review Skill

Applicability

• Platforms: iOS and Android
• React Native: 0.76+ (New Architecture interop assumed unless a checklist item says otherwise)

When to Use

• Reviewing a pull request or code change
• Auditing an existing screen or feature for issues
• Running a pre-merge quality gate
• Investigating the root cause of a production bug

This skill is the entry point for a full review. It does not restate every check — each concern lives in its own focused skill. Work through the relevant skills below, then report findings in the output format.

Prerequisites

• Access to the code being reviewed
• An understanding of the feature's intended behaviour
• Knowledge of which data is user-sensitive or business-critical

How to Conduct a Review

Review each concern using its focused skill. Each link resolves to that skill's checklist, examples, and pitfalls — apply those there rather than duplicating them here.

| Concern | Focused Skill | What it covers | |---|---|---| | Architecture | [architecture](../architecture/SKILL.md) | Folder structure, navigation, deep linking, rendering safety, error boundaries, code quality | | Performance | [performance](../performance/SKILL.md) | Re-renders, list virtualisation, images, animations, memory and resource cleanup | | Accessibility | [accessibility](../accessibility/SKILL.md) | Labels, roles, touch targets, focus order, colour, dynamic type | | Testing | [testing](../testing/SKILL.md) | Unit, component, and E2E coverage; edge cases; test quality | | State & Data | [state-and-data](../state-and-data/SKILL.md) | Server state, caching, offline, network transitions, transactions, loading/empty/error states | | Security | [security](../security/SKILL.md) | Secrets, secure storage, transport, PII in logs, deep-link validation, privacy compliance |

For each concern: open the focused skill, apply its checklist to the code under review, and collect anything that fails into the output format below. If a concern does not apply to the change, note it as not applicable.

Output Format

Structure every review as:

1. Summary — overall assessment (solid / needs work / significant issues)
2. Must-Fix — crashes, data loss, security or compliance issues (block merge)
3. Should-Fix — performance problems, architecture violations, missing tests
4. Consider — alternative approaches, future-proofing, style
5. What's Good — well-implemented patterns worth reinforcing

Pitfalls

• Reviewing only the happy path — the focused skills exist to make sure offline, error, empty, and accessibility states get checked too.
• Restating a focused skill's checklist here instead of linking to it causes drift; keep the detail in one place.
• Skipping the "What's Good" section removes the positive reinforcement that makes reviews land well.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: Neha
• Source: Neha/rn-developer-skills
• License: MIT

Install and usage instructions live in the source repository linked above.