claude-code-buddy

🥈Silver

Claude Code Buddy is an MCP server that enhances Claude Code with project awareness, memory, and task routing. It benefits developers by providing context-aware responses and learning from feedback. It integrates with Claude Code and can be applied to improve coding workflows and project management.

730Updated 1mo ago

Intermediate30min to implementautomation

Saves ~240 min per use

Quick InstallView Source

git clone https://github.com/PCIRCLE-AI/claude-code-buddy.git

Works with:

Claude

Overview

About This Skill

MeMesh LLM Memory provides persistent, searchable memory for Claude Code and MCP-compatible coding agents. It solves the problem of agents forgetting context between sessions by storing architectural decisions, bug fixes, failed tests, and lessons in a local SQLite database. The system uses FTS5 full-text search for fast retrieval without embeddings or LLM calls, achieving 95.40% recall on the LongMemEval-S benchmark. Installation is straightforward: Claude Code users can install as a plugin, while terminal users can install globally via npm. Both paths write to the same knowledge graph database, allowing memories captured in chat to appear in shell commands and vice versa.

How to Use

["Install the Claude Code Buddy MCP server in your Claude Code environment using the command: `claude install mcp://claude-code-buddy`","Load your project context by running: `claude load-project --path /path/to/your/project` to enable project-aware responses","Use the prompt template to frame your request, replacing [PLACEHOLDERS] with your specific needs (e.g., [PROJECT_NAME], [MODULE_NAME])","Review the MCP server's memory by asking: 'What prior feedback exists about [TOPIC]?' to leverage historical context","Iterate on the output by providing feedback: 'The suggested approach has a performance bottleneck at [SPECIFIC_LINE]. Can we optimize this?'"]

Use Cases

Preserve architectural decisions and design constraints across coding sessions

Auto-capture test failures and debugging solutions for future reference

Enable Claude Code agent to recall past project context without re-explanation

Share memories between Claude Code chat and terminal MCP clients like Cursor and Cline

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/PCIRCLE-AI/claude-code-buddy

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 Code Buddy MCP server to analyze the [PROJECT_NAME] codebase. Identify [SPECIFIC_TASK] such as refactoring [MODULE_NAME], fixing [BUG_TYPE], or implementing [FEATURE_NAME]. Provide a step-by-step plan with code snippets, trade-offs, and testing recommendations. Reference the project's existing patterns and suggest optimizations based on prior feedback stored in the MCP server's memory.

Example Output

For a fictional e-commerce platform called 'ShopFlow', I asked the Claude Code Buddy to refactor the payment processing module to improve security and maintainability. The MCP server analyzed the existing codebase, identified that the `PaymentService` class had 12 methods with 37% code duplication, and suggested consolidating payment validation logic into a new `PaymentValidator` class. It provided a 5-step migration plan with code snippets showing how to extract validation methods, update dependencies, and add comprehensive unit tests. The server also referenced prior feedback from the team's 'security review' session where they flagged inconsistent error handling in payment flows. It recommended adding OpenAPI schema validation and integrating with the new 'PaymentFraudDetection' microservice. The output included a diff of the proposed changes, estimated time savings of 15 developer hours, and a reminder to update the API documentation after implementation. The MCP server's memory suggested testing the changes with a canary deployment to 5% of users first, based on previous successful rollouts.