promptic

Promptic is an MCP (Model Context Protocol) server for prompt quality analysis, history management, and project version tracking. Connect it to Claude Code or any MCP-compatible client.

Features

• Prompt Analysis — Score prompts across 5 axes: clarity, completeness, specificity, context, and output format
• Version Tracking — Chain prompt versions (v1 → v2 → v3) and compare diffs
• History Management — Store, search, filter, and paginate prompt history
• Project Organization — Group prompts into named projects with snapshots
• Export / Import — Export as JSON, Markdown, or CSV; import Claude conversation logs
• Dashboard — Generate an HTML visualization with score trends and grade distribution

Tools (18 total)

| Category | Tools | |---|---| | Projects | list_projects, create_project, set_active_project, snapshot_project | | Analysis | analyze_prompt, compare_prompts, get_versions | | History | add_history_entry, get_history, query_history | | Export | visualize_project, export_project, import_project, import_claude_conversations | | Settings | get_settings, set_api_key, get_stats |

Installation

Claude Code / Cowork

npx @cppis/promptic

Or add to your MCP config (~/.claude.json / mcp.json):

{
  "mcpServers": {
    "promptic": {
      "command": "npx",
      "args": ["-y", "@cppis/promptic"]
    }
  }
}

Manual

npm install -g @cppis/promptic
promptic

Usage

Once connected, use the tools via natural language in Claude:

>> anz  (or >> 분석)        — Analyze the last prompt
>> deep                    — Deep analysis using Claude API
>> re   (or >> v2)         — Re-analyze with improvements

Analysis Modes

| Mode | Description | |---|---| | local | Fast rule-based scoring (default) | | api | Claude API-powered analysis | | deep | Full deep analysis with improvement suggestions |

Requirements

• Node.js >= 18
• Anthropic API key (optional — required for api / deep modes)

License

ISC