Agent skill

review-task

Tech lead review of task implementation. Use when user says "zkontroluj task", "review task", "zhodnot implementaci", or runs /review-task.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/review-task

SKILL.md

Review Task (Tech Lead Mode)

Perform a thorough tech lead review of task implementation - analyze code quality, verify against specs, check best practices from official documentation.

Usage

/review-task              # Deep review with comprehensive web research (default)
/review-task 02           # Review specific task
/review-task --quick      # Quick review - skip web research

Current State

Current branch: !git branch --show-current

Recent commits on this branch: !git log --oneline -10

Changed files vs main: !git diff --name-only main...HEAD 2>/dev/null || git diff --name-only HEAD~5

Process

Step 1: Identify Task to Review

Auto-detection order:

  1. If on feature branch phase-XX/task-YY-* → use that task
  2. If argument provided → use specified task number
  3. Find task with 🔵 in_progress status in current phase

Locate task file:

bash
# Find current phase from branch or recent work
PHASE_DIR=$(ls -d specification/phase-*/ | tail -1)
TASK_FILE=$(ls "$PHASE_DIR/tasks/task-${TASK_NUM}"*.md 2>/dev/null | head -1)

Step 2: Gather Context

Read these files in parallel:

  1. Task specification - the task .md file (scope, requirements)
  2. Phase specification - the phase.md file (objectives, related specs)
  3. Related high-level specs - linked specification documents
  4. Implementation files - all files changed for this task

Find changed files:

bash
# Files changed in this task's commits
git diff --name-only main...HEAD

# Or if on main, find by commit messages with task prefix [XX-YY]
git log --oneline --name-only --grep="\[${PHASE}-${TASK}\]" main | grep -v "^\w"

Step 3: Analyze Implementation

For each changed/created file, evaluate:

Code Quality Checklist

  • SOLID principles - Single responsibility, proper abstractions
  • Clean Architecture - Correct layer placement (Domain → Application → Infrastructure → API)
  • DDD patterns - Proper use of entities, value objects, aggregates, domain events
  • Error handling - Appropriate exceptions, validation, edge cases
  • Naming conventions - Clear, consistent, following project standards
  • Code duplication - No unnecessary repetition, proper abstractions
  • Dependency injection - Correct service registration and lifetimes

.NET Specific Checks

  • Async/await - Proper async patterns, no blocking calls
  • Nullable reference types - Proper null handling
  • IDisposable - Resources properly disposed
  • EF Core - Efficient queries, no N+1 problems
  • Configuration - Proper use of IOptions, no hardcoded values

Step 4: Verify Against Specifications

Compare implementation with task scope:

  1. Are all scope items implemented?
  2. Does implementation match the design in related specs?
  3. Are there any deviations that need justification?

Step 5: Web Research (Best Practices)

Use WebSearch and WebFetch to verify implementation against official docs and best practices.

Key research areas based on technologies used:

Technology What to verify
YARP Official configuration patterns, middleware order
MassTransit Consumer patterns, retry policies, outbox
EF Core Query patterns, migrations, concurrency
.NET Aspire Service defaults, health checks, telemetry
gRPC Proto best practices, error handling
Rate Limiting .NET 8+ built-in patterns

Search queries to use:

  • {technology} best practices site:learn.microsoft.com
  • {technology} official documentation
  • {pattern} .NET implementation guide

IMPORTANT: Always cite sources when recommending changes based on web research.

Step 6: Generate Review Report

Structure the output as follows:

markdown
# Task Review: [Task Name]

## Summary
[1-2 sentence overall assessment]

## Specification Compliance
| Scope Item | Status | Notes |
|------------|--------|-------|
| Item 1 | ✅/⚠️/❌ | ... |

## Strengths 💪
- [What was done well]
- [Good patterns used]
- [Proper architecture decisions]

## Issues Found 🔍

### Critical (must fix)
- [ ] **[Issue title]** in `file:line`
  - Problem: ...
  - Suggestion: ...
  - Reference: [link to docs]

### Improvements (should fix)
- [ ] **[Issue title]** in `file:line`
  - Problem: ...
  - Suggestion: ...

### Nitpicks (optional)
- [ ] ...

## Best Practices Verification
| Area | Status | Source |
|------|--------|--------|
| [Pattern] | ✅/⚠️ | [MS Docs link] |

## Recommendations
1. [Prioritized action items]
2. ...

## Questions for Developer
- [Clarifying questions if any]

Arguments

  • $ARGUMENTS - Task number or flags
    • 02 - specific task number
    • --quick - skip web research, just code review
    • --trace <correlation-id> - aggregate logs across services by CorrelationId

Review Depth Levels

Flag Code Analysis Spec Check Web Research
(default) Full Yes Comprehensive
--quick Full Yes Skip

Log Trace Mode (--trace)

When debugging issues during review, use the --trace flag to aggregate logs across all services by CorrelationId.

Usage:

bash
/review-task --trace 228617a4-175a-4384-a8e2-ade916a78c3f

What it does:

  1. Searches all service log files (gateway, order, product, notification, analytics)
  2. Finds all entries matching the CorrelationId
  3. Sorts entries chronologically across services
  4. Displays a unified trace of the request flow

Log format (Serilog):

[{Timestamp:HH:mm:ss} {Level:u3}] [{CorrelationId}] {Message:lj}

Example output:

[08:54:42 INF] [228617a4-...] Proxying to http://localhost/api/orders
[08:54:42 INF] [228617a4-...] Creating order for customer...
[08:54:42 INF] [228617a4-...] Reserving stock via gRPC...
[08:54:43 ERR] [228617a4-...] gRPC error in ReserveStock

Tool script:

bash
./tools/e2e-test/trace-correlation.sh <correlation-id> [--all-logs] [--json]

Options:

  • --all-logs - Search all log files, not just latest per service
  • --json - Output as JSON for programmatic use

Output Example

# Task Review: YARP Configuration

## Summary
Solid implementation of YARP reverse proxy with good route configuration.
Minor improvements needed in error handling and health check integration.

## Specification Compliance
| Scope Item | Status | Notes |
|------------|--------|-------|
| Configure YARP routes | ✅ | All routes properly defined |
| Add health checks | ⚠️ | Missing destination health checks |
| Load from config | ✅ | Using appsettings.json correctly |

## Strengths 💪
- Clean separation of route configurations
- Proper use of typed configuration binding
- Good middleware ordering

## Issues Found 🔍

### Critical (must fix)
None

### Improvements (should fix)
- [ ] **Missing destination health checks** in `Program.cs:45`
  - Problem: YARP health probing not configured
  - Suggestion: Add `.LoadFromConfig().AddHealthChecks()`
  - Reference: https://learn.microsoft.com/aspnet/core/host-and-deploy/health-checks

### Nitpicks (optional)
- [ ] Consider extracting route names to constants

## Best Practices Verification
| Area | Status | Source |
|------|--------|--------|
| YARP config | ✅ | MS Docs - YARP Configuration |
| Middleware order | ✅ | ASP.NET Core fundamentals |

## Recommendations
1. Add health checks for backend services
2. Consider adding request/response logging middleware

Integration

This skill works with the task workflow:

  1. /start-task XX - begin working
  2. /commit - commit changes
  3. /review-task - get tech lead feedback before finishing
  4. /finish-task - complete after addressing review

Safety Rules

  1. NEVER modify any files - this is a read-only review
  2. ALWAYS cite sources for recommendations from web research
  3. ALWAYS be constructive - balance criticism with praise
  4. ALWAYS prioritize issues (critical vs nice-to-have)
  5. If unsure about a pattern, research it before flagging as issue

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results