Skip to content

REVIEWS.md

Latest commit

 

History

History
48 lines (35 loc) · 2.68 KB

REVIEWS.md

File metadata and controls

48 lines (35 loc) · 2.68 KB

Code Review — Agent Instructions

Instructions for LLM agents. Keep edits minimal (headers + bullets). Use /agents-md skill when editing.

Applies to all review agents (human-assisted, Bugbot, code-review skill). Review bots should verify changes align with the nested AGENTS.md files: Sources/AGENTS.md, Tests/AGENTS.md, scripts/AGENTS.md, .github/AGENTS.md, Samples/AGENTS.md, develop-docs/AGENTS.md.

Review Priorities

In order of importance:

  1. Thread safety — SDK runs on arbitrary queues; check for data races, unprotected shared state, missing synchronization
  2. Memory management — retain cycles, leaks in ObjC code, strong reference cycles in closures/blocks
  3. Public API surface — backward compatibility, nullability annotations, NS_SWIFT_NAME correctness, SPI vs public visibility
  4. Cross-platform correctness — #if os(...) guards, @available annotations, platform-specific imports
  5. Error handling — silent failures, swallowed errors, missing fallback paths (the SDK must never crash the host app)
  6. Performance — hot path allocations, unnecessary main thread work, serialization overhead

SDK-Specific Concerns

  • Never crash the host app — all public entry points must be defensive
  • SentrySDK.start() can be called from any thread — initialization must be thread-safe
  • Swizzling — must be idempotent and check for prior swizzling
  • C/C++ code (SentryCrash) — buffer overflows, null pointer dereferences, signal safety
  • Session Replay — privacy-sensitive; verify redaction/masking logic
  • Envelope serialization — correct byte ordering, length prefixes, JSON encoding

Conventions to Enforce

  • Test names: test<Function>_when<Condition>_should<Expected>()
  • New/changed code must have corresponding tests
  • guard case over if case for pattern matching in tests
  • XCTUnwrap + element(at:) for safe array access (not direct subscript)
  • File renames preserve git history (git mv)
  • No AI assistant references in commits or PR descriptions
  • ObjC uses [[Class alloc] init], not [Class new]

What NOT to Flag

  • Style/formatting — handled by pre-commit hooks (SwiftLint, clang-format, dprint)
  • Test verbosity — tests follow DAMP (not DRY); duplicate test code is acceptable
  • // -- Arrange -- / // -- Act -- / // -- Assert -- — required test pattern
  • Conventional commit format — validated by CI

PR Description Expectations

  • Non-changelog changes include #skip-changelog
  • Breaking changes clearly documented
  • Untestable error paths documented