Claude Code Plugins

Community-maintained marketplace

Agent Skills

implementation-phase

@marcusgoll/robinhood-algo-trading-bot

1

0

Standard Operating Procedure for /implement phase. TDD workflow, anti-duplication checks, task execution, and continuous testing.

Install Skill

1Download skill

2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name	implementation-phase
description	Standard Operating Procedure for /implement phase. TDD workflow, anti-duplication checks, task execution, and continuous testing.
allowed-tools	Read, Write, Edit, Grep, Glob, Bash

Implementation Phase: Quick Reference

Purpose: Execute tasks from tasks.md using Test-Driven Development, preventing code duplication, and maintaining high quality.

Phase Overview

Inputs:

specs/NNN-slug/tasks.md - Task breakdown (20-30 tasks)
specs/NNN-slug/plan.md - Implementation plan with reuse strategy

Outputs:

Implemented code (models, services, APIs, UI components)
Test suites (unit, integration, E2E)
Updated tasks.md and workflow-state.yaml

Expected duration: 2-10 days (varies by complexity)

Quick Start Checklist

Before you begin:

Tasks phase completed (tasks.md exists with 20-30 tasks)
Plan phase completed (plan.md exists)
Development environment set up
Test framework configured
Git working tree clean

Core workflow:

✅ Load Tech Stack Constraints - Prevent hallucinated tech choices
✅ Review Task Dependencies - Identify parallel work opportunities
✅ Execute Tasks Using TDD - RED → GREEN → REFACTOR
✅ Update Task Status - Keep NOTES.md current
✅ Run Anti-Duplication Checks - Search before writing
✅ Continuous Testing - Test after each task triplet
✅ Commit Implementation - Small, frequent commits

Detailed Resources

🎯 Core Workflow

TDD Workflow - RED → GREEN → REFACTOR cycle, test-first discipline
Task Batching - Parallel execution strategy, dependency analysis
Task Tracking - NOTES.md updates, velocity tracking

🛡️ Quality Gates

Tech Stack Validation - Load constraints from tech-stack.md
Anti-Duplication Checks - Search patterns, DRY enforcement
Continuous Testing - Test cadence, coverage requirements

🚧 Advanced Topics

Handling Blocked Tasks - Escalation strategies, workarounds
Integration Testing - Multi-component tests
UI Component Testing - React Testing Library, accessibility
E2E Testing - Playwright/Cypress patterns

📋 Reference

Common Mistakes - Anti-patterns to avoid
Best Practices - Proven patterns from production
Code Review Checklist - Pre-commit validation
Troubleshooting Guide - Common blockers and fixes

Completion Criteria

Required:

All tasks completed (or blocked tasks documented)
Test coverage ≥80% (unit + integration)
All tests passing (CI green)
No code duplication (DRY violations <3)
Code review checklist passed
Git commits made with descriptive messages

Optional (if applicable):

UI components have accessibility tests (WCAG 2.1 AA)
E2E tests cover critical user flows
Performance benchmarks met (API <200ms, page load <2s)

Next Phase

After implementation complete: → /optimize - Code review, performance validation, production readiness

See also:

reference.md - Comprehensive implementation guide (full text)
examples.md - Good vs bad implementation examples
scripts/batch-validator.sh - Batch task validation