vibemaster-mcp
v7.1.0
Published
The only MCP that makes Claude think like a senior engineer — memory, GSD planning, auto-review, quality scoring, GitHub Action CI. 58 tools.
Downloads
963
Maintainers
Readme
VibemasterMCP
A Model Context Protocol server that gives Claude deep code analysis superpowers: 60-rule bug detection, dataflow taint analysis, auto-patching, refactor suggestions, complexity tracking, and workflow orchestration.
✨ What It Does
VibemasterMCP transforms Claude from a smart assistant into a senior engineer partner. It adds persistent memory, enforces code quality standards, catches bugs before they reach you, and orchestrates multi-step workflows automatically.
Core Capabilities:
- 60-Rule Bug Detection — Static analysis for async errors, type safety, React patterns, security vulnerabilities, performance issues
- Dataflow Taint Analysis — Trace user input through your code to dangerous sinks (SQL injection, XSS, RCE)
- Auto-Patcher — Automatically applies safe mechanical fixes (any→unknown, var→const, empty catch handlers)
- Complexity Tracking — Grade files A-F with historical trends
- Multi-File Architecture Analysis — Detect circular dependencies, bottlenecks, orphaned code
- Quality Scoring — Every output gets 0-100 with specific feedback
- Persistent Memory — Remembers your project conventions across sessions
🚀 Quick Start
Installation
npm install
npm run buildAdd to Claude Desktop
Add this to your claude_desktop_config.json:
{
"mcpServers": {
"vibemaster": {
"command": "node",
"args": ["/absolute/path/to/VibemasterMCP/dist/index.js"]
}
}
}Config locations:
- macOS/Linux:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
First Use
In Claude, type:
run health_checkVerify all subsystems are operational.
🛠️ Tools Reference
| Tool | Description | Key Parameters |
|------|-------------|----------------|
| debug_session | Full bug analysis on code snippets | input, mode |
| run_workflow | Multi-step analysis workflow | workflow_id, code |
| auto_patch | Apply safe mechanical fixes | code, preview_only |
| diff_analyze | Show only bugs introduced by changes | original, changed |
| dataflow_analyze | Trace tainted data to dangerous sinks | code |
| complexity_check | Grade file complexity A-F with trend | code, file_path |
| analyze_project | Multi-file architecture health check | files (map of path→content) |
| refactor_suggest | Concrete refactoring opportunities | code, file_path |
| generate_tests | Generate Vitest regression tests | code, module_name |
| analyze_prompt | Score AI prompt quality + injection detection | prompt |
| quality_report | Overall code quality score 0-100 | content, language |
| code_review | Line-level PR-style review comments | code, file_path |
| call_graph | Map function call relationships | code, target_function |
| session_report | Full session analysis summary | — |
| session_patterns | Detect recurring anti-patterns | — |
| health_check | Verify all subsystems operational | — |
| self_test | Run end-to-end smoke tests | — |
| workflow_guide | Get guidance on which tool to use | — |
🔄 Workflows
VibemasterMCP includes 5 built-in workflows that chain multiple tools:
| Workflow | Description |
|----------|-------------|
| full_review | Complete analysis: bugs, quality, complexity, tests, smart explanations |
| security_audit | Taint flow analysis + injection vulnerability checks |
| pre_commit | Fast check before committing — catch issues early |
| onboarding | First-time analysis of a new file or module |
| fix_and_verify | Apply fixes, generate tests, verify complexity |
Usage:
run_workflow workflow_id=security_audit code="<your code>"🔗 Git Hook Integration
Create a pre-commit hook to remind yourself to run the pre_commit workflow:
#!/bin/sh
# .git/hooks/pre-commit
echo "Reminder: Run the 'pre_commit' workflow in Claude before committing."Or with Husky:
npx husky add .husky/pre-commit "echo 'Reminder: run pre_commit workflow in Claude before committing'"📁 Architecture
src/
├── analysis/
│ ├── bug-detector.ts # 60-rule static bug detection
│ ├── dataflow-analyzer.ts # Taint flow analysis
│ ├── complexity-tracker.ts # Complexity scoring + history
│ ├── call-graph.ts # Function call relationships
│ ├── diff-analyzer.ts # Changed code analysis
│ ├── project-analyzer.ts # Multi-file architecture health
│ ├── refactor-suggester.ts # Concrete refactoring ideas
│ ├── session-reporter.ts # Session summary generation
│ └── session-pattern-tracker.ts # Anti-pattern detection
├── core/
│ ├── tool-validator.ts # Input validation for all tools
│ ├── tool-cache.ts # Tool result caching
│ ├── error-boundary.ts # Error handling + categorization
│ ├── health-check.ts # Subsystem diagnostics
│ ├── workflow-engine.ts # Multi-tool workflow orchestration
│ └── output-limiter.ts # Response size management
├── detectors/
│ ├── auto-patcher.ts # Safe mechanical code fixes
│ ├── mcp-detector.ts # Find installed MCPs
│ ├── stack-detector.ts # Detect project tech stack
│ └── runtime-bug-detector.ts # Runtime error patterns
├── orchestration/
│ ├── orchestrator.ts # Main MCP coordination
│ └── role-engine.ts # Agent role handoff
├── quality/
│ ├── code-reviewer.ts # Line-level review generation
│ ├── quality-scorer.ts # 0-100 quality scoring
│ └── response-composer.ts # Unified output formatting
├── security/
│ ├── secret-guard.ts # Block secrets in output
│ ├── permission-guard.ts # Role-based permissions
│ └── rate-limiter.ts # Request rate limiting
├── memory/
│ └── memory-manager.ts # Persistent project memory
└── server-core.ts # Tool definitions + handlers🏗️ Development
npm run build # Compile TypeScript to dist/
npm run dev # Watch mode for development
npm run start # Run the MCP server
npm test # Run vitest test suite📈 Version History
| Version | Key Addition | |---------|-------------| | v2.2.7 | Initial release | | v2.3.0 | RuntimeBugDetector, MCPReliability | | v2.4.0 | TestGenerator, PromptAnalyzer, SessionReporter | | v2.5.0 | ComplexityTracker, CallGraph, SmartExplainer | | v2.6.0 | ToolValidator, ToolCache, ErrorBoundary, HealthCheck | | v2.7.0 | WorkflowEngine, DiffAnalyzer, OutputLimiter | | v2.8.0 | DataflowAnalyzer, SessionPatternTracker | | v2.9.0 | AutoPatcher, RefactorSuggester | | v3.0.0 | 60 detection rules (+25: async, React, security, perf) | | v3.1.0 | ProjectAnalyzer (multi-file arch), README |
License: MIT © SnaiilyDevelopment
GitHub: https://github.com/SnaiilyDevelopment/VibemasterMCP
