Your AI agent,
supercharged
One command. 13 tools. Your AI coding agent gets 1,500+ battle-tested prompts, smart enhancement, workflow chains, config generation, and real-time code quality scoring with auto-fix — directly inside your IDE.
Works offline in Claude Code, Cursor, Windsurf, and GitHub Copilot. No API keys needed.
npx promptrr-mcp-serverPrompt Tools
6 tools
Your agent enhances vague ideas into production-ready prompts, fixes bad AI output, searches 2,500+ curated prompts, builds multi-step chains, and generates config files for 8 platforms.
Drift Detector
7 tools
Scores your code 0-100 against conventions in CLAUDE.md, .cursorrules, or any AI config file. Full file analysis, quick scans, auto-fix violations, track drift history, set baselines, and compare trends — all offline, zero AI calls.
Prompt Tools
Enhance, fix, search, or generate — your agent picks the right tool automatically.
Turn a vague idea into a production-ready prompt
Analyze bad AI code and get a better prompt
Search 2,500+ curated prompts by keyword
Generate from pre-built templates with guided inputs
Multi-step prompt chains for complex features
Generate CLAUDE.md, .cursorrules, copilot-instructions
Drift Detector
Scores code 0-100, auto-fixes violations, tracks history & baselines. Zero AI calls.
Full analysis of code files against conventions from CLAUDE.md, .cursorrules, etc.
Instant single-snippet scan with default conventions
Extract coding rules from any AI config file
Before/after comparison — see if your changes improved quality
Auto-fix common violations — var→const, console removal, type imports, and more
Track drift scores over time — record entries, view trends, and monitor progress
Set a quality baseline and compare future scans against it
Add to your IDE
Paste the config into your IDE's MCP configuration file. Your AI agent will discover all 13 tools, 4 resources, and 4 prompts automatically.