Metadata-Version: 2.4
Name: claudeboost-mcp
Version: 0.6.1
Summary: MCP server that enhances prompts for Claude Code using enterprise playbook rules
Project-URL: Homepage, https://claudeboost.com
Project-URL: Repository, https://github.com/FarhadS08/claudeboost
Project-URL: Documentation, https://github.com/FarhadS08/claudeboost#readme
Author-email: ClaudeBoost <hello@claudeboost.com>
License-Expression: MIT
Keywords: ai,claude,claude-code,mcp,prompt-engineering
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Software Development :: Libraries
Requires-Python: >=3.10
Requires-Dist: anthropic>=0.30.0
Requires-Dist: mcp>=1.0.0
Description-Content-Type: text/markdown

# ClaudeBoost MCP

> Supercharge your Claude Code prompts with enterprise-grade enhancement.

ClaudeBoost is an MCP server that transforms vague prompts into structured, enterprise-quality instructions. It classifies your prompt's domain, scores it across 6 dimensions, and rewrites it using playbook-powered rules.

## Quick Start

### 1. Install

```bash
pip install claudeboost-mcp
```

### 2. Add to Claude Code

```bash
claude mcp add claudeboost -- claudeboost-mcp
```

Or add manually to `~/.claude/mcp_settings.json`:
```json
{
  "mcpServers": {
    "claudeboost": {
      "command": "claudeboost-mcp",
      "env": {
        "ANTHROPIC_API_KEY": "your-key-here"
      }
    }
  }
}
```

### 3. Set your Anthropic API key

```bash
export ANTHROPIC_API_KEY="sk-ant-..."
```

### 4. Use it

```
/boost build me an ETL pipeline
```

Or just type any prompt — ClaudeBoost auto-boosts all task prompts.

## Features

- **7 Domain Detection** — Data Science, Data Engineering, Business Analytics, General Coding, Documentation, DevOps, Other
- **6-Dimension Scoring** — Specificity, Verification, Context, Constraints, Structure, Output (1-5 each, /30 total)
- **3 Boost Levels** — Light (clarify only), Medium (balanced), Full (enterprise playbook)
- **RLHF Feedback Loop** — Rate boosts and leave feedback → improves future enhancements
- **Web Dashboard** — View history, stats, score histograms at [claudeboost.com](https://claudeboost.com)

## Commands

| Command | Description |
|---------|-------------|
| `/boost <prompt>` | Boost a specific prompt |
| `/boost --login` | Sign in (opens browser) |
| `/boost --logout` | Sign out |
| `/boost-settings --level light\|medium\|full` | Change boost intensity |
| `/boost-settings --auto true\|false` | Toggle auto-boost |
| `/boost-help` | Show all commands |
| `--raw` | Add to end of prompt to skip boost |

## Requirements

- Python 3.10+
- Anthropic API key
- Claude Code CLI

## License

MIT
