Review Code

Review Code

Step 1 — Assess PR scope

# Count changed files and lines
git diff --stat HEAD~1

• If >500 lines changed: flag for splitting before detailed review
• If >20 files changed: flag for splitting before detailed review

Step 2 — Check PR metadata

Verify:

• ☐ PR title follows conventional commits (type(scope?): description (#issue) — NO [context] prefix; [context] is for commit messages only)
• ☐ PR description explains WHAT and WHY
• ☐ Linked issue/ticket exists

Step 3 — Check build

mvn spotless:check -pl openaev-api
mvn compile -pl openaev-api -q

If build fails: stop review, report build failure.

For frontend changes:

cd openaev-front && yarn check-ts && yarn lint

Step 4 — Review architecture alignment

# Check for entity exposure in API layer (should use DTOs)
grep -rn "import io.openaev.database.model" openaev-api/src/main/java/io/openaev/api/ openaev-api/src/main/java/io/openaev/rest/ --include="*.java" | grep -v "Action\|ResourceType\|Capability\|Filters\|Grant"

Flag any direct entity usage in controllers or API responses (should use Output records + Mapper).

# Check for repository injection in controllers (should go through service)
grep -rn "Repository" openaev-api/src/main/java/io/openaev/api/ openaev-api/src/main/java/io/openaev/rest/ --include="*.java" | grep -v "test\|Test"

Step 5 — Review code quality

# System.out.println (should use @Slf4j)
grep -rn "System.out\|System.err\|printStackTrace" --include="*.java" openaev-api/src/main/java/

# jakarta.transaction.Transactional (should use Spring's)
grep -rn "jakarta.transaction.Transactional" --include="*.java" openaev-api/src/main/java/

# New code in deprecated module
git diff --name-only HEAD~1 | grep "openaev-framework"

# New code in legacy rest/ package (should be in api/)
git diff --name-only HEAD~1 | grep "io/openaev/rest/" | grep -v "test"

Step 6 — Check test coverage

# Are there test files for the changed production files?
for f in $(git diff --name-only HEAD~1 | grep "src/main/java" | grep -v "migration"); do
  testfile=$(echo "$f" | sed 's|src/main/java|src/test/java|' | sed 's|\.java$|Test.java|')
  if [ ! -f "$testfile" ]; then
    echo "⚠️ Missing test: $testfile"
  fi
done

Step 7 — Determine delegation

Based on changed files, determine if specialized agents should run:

# Security signals
grep -rn "AccessControl\|@Filter\|Capability\|Permission\|nativeQuery" --include="*.java" $(git diff --name-only HEAD~1) 2>/dev/null | head -10

# Performance signals
grep -rn "OneToMany\|ManyToMany\|FetchType\|findAll\|Pageable" --include="*.java" $(git diff --name-only HEAD~1) 2>/dev/null | head -10

# Tenancy signals (v1 @Filter + v2 TxCtx/active-tables)
grep -rn "TenantBase\|tenant_id\|TenantContext\|TxCtx\|active-tables\|TenantScopedTransaction\|RequireTenantSelector\|can_access_tenant" --include="*.java" $(git diff --name-only HEAD~1) 2>/dev/null | head -10

# Frontend signals
git diff --name-only HEAD~1 | grep -E "\.tsx$|\.ts$" | head -10

Step 8 — Check common anti-patterns (learned from reviews)

Apply these checks based on past review feedback:

• Root cause vs workaround: If a bug is backend-originated, don't add frontend workarounds (onError handlers, fallback states). Fix the root cause in the correct layer.
• String/value duplication: When introducing new methods that share data with existing methods (e.g., logo filenames, config keys), extract the shared value to a private method or constant — never compute the same formatted string in multiple places.
• Non-critical operations: Startup operations that interact with external services (MinIO, S3, etc.) for non-critical assets (logos, thumbnails) should be best-effort: wrap in try-catch with log.warn so failures don't block application startup.
• Test proportionality: Don't require tests for small mechanical changes (1-line additions, parameter threading). Tests should be proportionate to the risk and complexity of the change.
• Pre-existing issues: Don't fix unrelated pre-existing issues (e.g., alt text, parameter naming) in a bug fix PR. Track them as follow-up.

Step 9 — Compile review

Generate the Code Review Summary following the output format defined in code-reviewer.agent.md.