Automatically synchronizes prompt templates from GitHub repositories and includes specialized agents for GitHub workflows such as pull request reviews and repository exploration.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@MCP Prompt Template Selectorfind the best template for reviewing github pull requests"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP Prompt Template Selector
MCP server that automatically selects and generates specialized prompts from a knowledge base of 66+ Claude Code system prompts. Uses LLM-powered semantic search to find the best matching template and adapts it to your specific task.
Features
Smart Template Selection — Semantic search across 66+ prompt templates using Claude Haiku
Adaptive Generation — Returns original template when it fits, customizes when adaptation is needed
Custom Templates — Add your own templates with priority over built-in ones
Dual Transport — Works via stdio (local) or HTTP (server deployment)
Zero Config — Templates auto-sync from claude-code-system-prompts
Requirements
Bun v1.0+
Anthropic API key
Installation
Quick Start
Connect to Claude Code (recommended)
Add MCP server to Claude Code via command line:
Verify connection:
Then in Claude Code use /mcp to see available tools.
Alternative: Run as HTTP server
Usage
Once connected, you have 4 tools available in Claude Code:
list_templates
Browse available templates:
select_template
Find best matching template for your task:
generate_prompt
Generate a ready-to-use prompt:
sync_templates
Update templates from GitHub:
Examples
Example 1: Code Review Agent
Request:
Result: Returns the original agent-prompt-review-pr-slash-command template as-is, since it's a perfect match.
Example 2: Custom Task with Adaptation
Request:
Result: Adapts the agent-prompt-plan-mode-enhanced template with Django/async specifics:
Example 3: Listing Agent Templates
Request:
Result:
Template Categories
Category | Count | Examples |
| 28 | PR review, code exploration, security audit |
| 10 | Base behavior, plan mode, configuration |
| 19 | Bash, Read, Edit, Grep tool descriptions |
| 3 | Plan mode reminders |
| 5 | GitHub integration, specialized skills |
Custom Templates
Add your own templates to custom-templates/ directory:
metadata.json:
Custom templates are prioritized over built-in ones during selection.
API Reference
Tools
Tool | Description | Required Params |
| List/search templates | — |
| Find matching templates |
|
| Generate customized prompt |
|
| Sync from GitHub | — |
Parameters
list_templates:
category:agent|system|tool|reminder|skill|anysearch: Text search queryinclude_custom: Include custom templates (default: true)
select_template / generate_prompt:
task_description: What you need the prompt forprompt_type: Filter by categorycontext: Additional context for adaptationtarget_llm: Target model (default: Claude)max_results: Max templates to consider (default: 3)
sync_templates:
force_reindex: Re-index all templates (default: false)
Configuration
Environment Variables
Variable | Description | Default |
| Anthropic API key | — |
| HTTP server port | 3000 |
Claude Code Integration
Global config (~/.claude.json):
Project config (.mcp.json in project root):
Development
Project Structure
How It Works
Sync: Clones/pulls claude-code-system-prompts
Index: Uses Claude Haiku to extract metadata from each template
Select: Semantic search finds best matching templates for your task
Generate: Returns original or adapts template based on match quality
License
MIT
Credits
Templates from Piebald-AI/claude-code-system-prompts
Built with Model Context Protocol SDK