minipuft-marketplace

Claude Prompts MCP Server

Why

Stop copy-pasting prompts. This server turns your prompt library into a programmable engine.

• Version Control — Prompts are Markdown files in git. Track changes, review diffs, branch experiments.
• Hot Reload — Edit a template, run it immediately. No restarts.
• Structured Execution — Not just text. The server parses operators, injects methodology, enforces quality gates, renders the final prompt.

Quick Start

MCP clients launch the server automatically—you just configure and connect.

Option 1: NPM (Fastest)

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "claude-prompts": {
      "command": "npx",
      "args": ["-y", "claude-prompts@latest"]
    }
  }
}

Restart Claude Desktop. Test with: prompt_manager(action: "list")

That's it. The client handles the rest.

---

Option 1b: NPM with Custom Prompts

Want your own prompts without cloning the repo? Create a workspace:

npx claude-prompts --init=~/my-prompts

This creates a workspace with starter prompts. Then point Claude Desktop to it:

{
  "mcpServers": {
    "claude-prompts": {
      "command": "npx",
      "args": ["-y", "claude-prompts@latest"],
      "env": {
        "MCP_WORKSPACE": "/home/YOUR_USERNAME/my-prompts"
      }
    }
  }
}

Restart Claude Desktop. Your prompts are now hot-reloadable—edit them directly, or ask Claude to update them:

User: "Make the quick_review prompt also check for TypeScript errors"
Claude: prompt_manager(action:"update", id:"quick_review", ...)  # Updates automatically

See the server README for all configuration options.

---

Option 2: From Source (For Customization)

Clone if you want to edit prompts, create custom frameworks, gates or contribute:

git clone https://github.com/minipuft/claude-prompts-mcp.git
cd claude-prompts-mcp/server
npm install && npm run build

Then configure Claude Desktop to use your local build:

Windows:

{
  "mcpServers": {
    "claude-prompts": {
      "command": "node",
      "args": ["C:\\path\\to\\claude-prompts-mcp\\server\\dist\\index.js"]
    }
  }
}

Mac/Linux:

{
  "mcpServers": {
    "claude-prompts": {
      "command": "node",
      "args": ["/path/to/claude-prompts-mcp/server/dist/index.js"]
    }
  }
}

---

Verify It Works

Restart Claude Desktop. In the input bar, type:

prompt_manager list

How It Works

Not a static file reader. It's a render pipeline with a feedback loop:

%%{init: {'theme': 'neutral', 'themeVariables': {'background':'#0b1224','primaryColor':'#e2e8f0','primaryBorderColor':'#1f2937','primaryTextColor':'#0f172a','lineColor':'#94a3b8','fontFamily':'"DM Sans","Segoe UI",sans-serif','fontSize':'14px','edgeLabelBackground':'#0b1224'}}}%%
flowchart TB
    classDef actor fill:#0f172a,stroke:#cbd5e1,stroke-width:1.5px,color:#f8fafc;
    classDef server fill:#111827,stroke:#fbbf24,stroke-width:1.8px,color:#f8fafc;
    classDef process fill:#e2e8f0,stroke:#1f2937,stroke-width:1.6px,color:#0f172a;
    classDef client fill:#f4d0ff,stroke:#a855f7,stroke-width:1.6px,color:#2e1065;
    classDef clientbg fill:#1a0a24,stroke:#a855f7,stroke-width:1.8px,color:#f8fafc;
    classDef decision fill:#fef3c7,stroke:#f59e0b,stroke-width:1.6px,color:#78350f;

    linkStyle default stroke:#94a3b8,stroke-width:2px

    User["1. User sends command"]:::actor
    Example[">>analyze @CAGEERF :: 'cite sources'"]:::actor
    User --> Example --> Parse

    subgraph Server["MCP Server"]
        direction TB
        Parse["2. Parse operators"]:::process
        Inject["3. Inject framework + gates"]:::process
        Render["4. Render prompt"]:::process
        Decide{"6. Route verdict"}:::decision
        Parse --> Inject --> Render
    end
    Server:::server

    subgraph Client["Claude (Client)"]
        direction TB
        Execute["5. Run prompt + check gates"]:::client
    end
    Client:::clientbg

    Render -->|"Prompt with gate criteria"| Execute
    Execute -->|"Verdict + output"| Decide