claude-research-plan-implement

🥈Silver

Structured AI development framework for Claude Code. Research, plan, and implement workflows with parallel agents, persistent context, and session management. Benefits operations teams by streamlining software development processes.

10150Updated 1mo ago

Intermediate30min to implementautomation

Saves ~240 min per use

Quick InstallView Source

git clone https://github.com/brilliantconsultingdev/claude-research-plan-implement.git

Works with:

Claude

Overview

About This Skill

How to Use

1. **Initialize the session**: Start a new Claude session with `claude --session research-plan-implement` and enable persistent context with `session save` after each major step. 2. **Launch parallel agents**: Use `claude --agents research,planner,implementer` to create specialized agents for each phase. Assign the research agent to gather information while the planner creates the architecture. 3. **Maintain context**: After each phase, use `session summarize` to capture decisions and `session log` to document the rationale behind technical choices. This ensures continuity between phases. 4. **Iterate with feedback**: After implementation, use `claude --review` to analyze the codebase against the original plan. Generate a diff report highlighting deviations and their business impact. 5. **Document outcomes**: Use `session export` to create a final report combining research findings, implementation details, and testing results for stakeholder review. Tip: For complex projects, break the implementation into 2-week sprints with checkpoints at each phase. Use the planner agent to create detailed task breakdowns with estimated hours and dependencies.

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/brilliantconsultingdev/claude-research-plan-implement

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Use the claude-research-plan-implement framework to develop a [PROJECT_NAME] feature for [TECH_STACK]. First, research existing solutions and best practices for [SPECIFIC_TASK] in [RELEVANT_DOMAIN]. Then, create a detailed implementation plan covering architecture, dependencies, and risk mitigation. Finally, implement the feature in a modular way, ensuring test coverage and documentation. Use parallel agents for research and implementation where beneficial. Maintain session context throughout the process and document all decisions in a persistent log.

Example Output

### Research Phase
I analyzed 12 GitHub repositories and 5 technical blogs to identify patterns for implementing real-time data synchronization in Next.js applications. The research revealed that 70% of production-grade solutions use WebSocket-based approaches with Redis pub/sub for scalability. I documented key findings in a research log, including trade-offs between server-sent events and GraphQL subscriptions. The most promising pattern emerged from the Vercel team's experimental real-time dashboard implementation, which uses a hybrid approach with optimistic UI updates and conflict resolution strategies.

### Planning Phase
I created a modular architecture with three core components: 1) A WebSocket handler service (300 lines) using the `ws` library with automatic reconnection logic, 2) A Redis adapter layer for pub/sub messaging with exponential backoff, and 3) A client-side SDK (200 lines) with TypeScript type safety and event batching. The implementation plan includes:
- Risk mitigation: Circuit breakers for WebSocket failures, fallback to polling every 5s
- Testing strategy: Unit tests for core logic, integration tests for WebSocket lifecycle, and load tests simulating 1000 concurrent connections
- Rollout plan: Feature flagged deployment with 10% traffic in week 1, gradual ramp to 100%

### Implementation Phase
Started parallel implementation with two agents: Agent A handled the WebSocket service and Redis integration, while Agent B focused on the client SDK and UI components. The WebSocket service now handles 99.8% of connections successfully (tested with 5000 concurrent users) with an average message latency of 45ms. The client SDK includes automatic reconnection with jitter (1s-3s delay) and message buffering during outages. All components are documented with JSDoc comments and include comprehensive error handling. The implementation exceeds the original requirements by adding automatic reconnection and message deduplication features not initially specified.