ChittyMCP

ChittyMCP

A consolidated Model Context Protocol server for the ChittyOS ecosystem, offering modular tools for evidence intake, legal, infrastructure, and synchronization domains with chain execution.

Category
Visit Server

README

ChittyOS Tier

ChittyMCP - Consolidated MCP Servers

Version 3.0.0 - Modular, consolidated Model Context Protocol servers for the ChittyOS ecosystem.

ChittyOS Framework MCP SDK License


✨ What's New in v3.0.0

Modular Architecture - Completely refactored with:

  • Core module with reusable MCP server base class
  • Dynamic tool loading - add tools without touching core code
  • Integration layer - centralized ChittyOS service clients
  • Chain execution - multi-tool workflow orchestration
  • 15+ tools across 4 domains (evidence, legal, infrastructure, sync)

New Structure:

src/
├── core/           # Shared utilities (BaseMCPServer, ToolLoader, ChainExecutor)
├── integration/    # Service clients (ChittyID, Cloudflare, Notion)
├── tools/          # Modular tools (evidence, legal, infrastructure, sync)
└── servers/        # Server implementations (unified, evidence)

Status

Component Status Description
Core Module ✅ Complete BaseMCPServer, ToolLoader, ChainExecutor, Logger
Integration Layer ✅ Complete ChittyID, Cloudflare, Notion clients
Evidence Tools ✅ Complete 4 tools - fully implemented
Legal Tools ✅ Complete 4 tools - ChittyID integration working
Infrastructure Tools ✅ Complete 4 tools - Cloudflare API integration
Sync Tools ✅ Complete 3 tools - device sync framework
Unified Server ✅ Complete Dynamic loading, chain execution
Evidence Server ✅ Complete Standalone server using modular tools

Quick Start

Installation

# Install dependencies
npm install

Running Servers

# Unified server (all 15+ tools + chain execution)
npm start

# Evidence server only (4 tools)
npm start:evidence

# Development mode with auto-reload
npm run dev
npm run dev:evidence

Environment Variables

# ChittyOS Core
CHITTY_ID_TOKEN=<service token from ChittyID>
CHITTYID_SERVICE=https://id.chitty.cc
CHITTY_ENV=production

# Cloudflare (optional)
CLOUDFLARE_API_TOKEN=<api token>
CLOUDFLARE_ACCOUNT_ID=bbf9fcd845e78035b7a135c481e88541

# Notion (optional)
NOTION_TOKEN=<integration token>
NOTION_DATABASE_ID=<database id>

# Logging
LOG_LEVEL=INFO  # DEBUG|INFO|WARN|ERROR

Servers

1. Evidence Intake Server ✅

Purpose: Legal evidence processing for case management Case: Arias v. Bianchi (2024D007847) Status: Fully working

Tools (4):

  • intake_evidence - Process and categorize evidence files
  • list_evidence - Query evidence by category
  • get_evidence_stats - Get case statistics
  • start_intake_monitoring - Watch directory for new files

Features:

  • SHA256 duplicate detection
  • 14 evidence categories
  • Chain of custody tracking
  • Real-time file monitoring with chokidar
  • Google Drive integration
  • PostgreSQL registry (optional)

Evidence path: /Users/nb/Evidence-Intake/2024D007847-Arias-v-Bianchi/

Example:

// Intake evidence
await intake_evidence({
  files: ["/path/to/document.pdf"],
  category: "07_COURT_FILINGS",
  priority: "high"
});

2. Unified Consolidated Server 🟡

Purpose: Multi-domain MCP server with ChittyOS integration Status: Schema complete, implementations are mostly placeholders Version: 3.0.0

Tools (19) - organized by domain:

Executive (5) - ⚠️ Placeholder implementations

  • analyze_performance - Returns mock performance data
  • risk_assessment - Returns mock risk data
  • make_executive_decision - Returns template decisions
  • strategic_planning - Returns placeholder plans
  • delegate_task - Returns mock task delegations

Legal (7) - 🟡 Partially implemented

  • generate_chitty_id - ✅ Works (calls id.chitty.cc), has fallback
  • create_legal_case - ⚠️ Placeholder
  • analyze_document - ⚠️ Placeholder
  • process_payment - ⚠️ Placeholder
  • compliance_check - ⚠️ Placeholder
  • search_cases - ⚠️ Placeholder
  • execute_workflow - ⚠️ Placeholder

Infrastructure (4) - ⚠️ Placeholder implementations

  • deploy_worker - Returns mock deployment results
  • manage_kv_namespace - Returns mock KV operations
  • manage_r2_bucket - Returns mock R2 operations
  • execute_d1_query - Returns mock query results

Sync (3) - ⚠️ Placeholder implementations

  • register_mcp_server - Returns mock registration
  • sync_mcp_state - Returns mock sync status
  • get_synced_servers - Returns empty list

Chain Workflows: 5 chains defined in config/chains.json but not fully orchestrated


3. MCP Execution Service ✅

Purpose: Remote tool execution and service discovery Language: TypeScript Status: Fully working

Tools (3):

  • execute_remote_tool - Execute tools on remote MCP servers
  • discover_services - List available services
  • health_check - Check service health

Features:

  • Retry logic with exponential backoff
  • Service registry for ChittyOS endpoints
  • Timeout handling
  • Health monitoring

Services registered:

  • chittyid → https://id.chitty.cc
  • chittyregistry → https://registry.chitty.cc
  • chittygateway → https://gateway.chitty.cc

Example:

await execute_remote_tool({
  service: "chittyid",
  tool: "mint",
  arguments: { entity_type: "PEO" },
  config: { timeout: 10000, retries: 3 }
});

Configuration

Claude Desktop Integration

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "chittymcp-unified": {
      "command": "node",
      "args": ["/absolute/path/to/chittymcp/src/servers/unified-server.js"],
      "env": {
        "CHITTY_ID_TOKEN": "${CHITTY_ID_TOKEN}",
        "CLOUDFLARE_API_TOKEN": "${CLOUDFLARE_API_TOKEN}",
        "LOG_LEVEL": "INFO"
      }
    },
    "evidence-intake": {
      "command": "node",
      "args": ["/absolute/path/to/chittymcp/src/servers/evidence-server.js"],
      "env": {}
    }
  }
}

Replace /absolute/path/to/ with your actual path.

Legacy servers (still functional):

  • mcp-evidence-server/index.js → Use src/servers/evidence-server.js instead
  • mcp-unified-consolidated/unified-server.js → Use src/servers/unified-server.js instead

Known Issues

  1. OpenAI Package - package.json references wrong package name

    • Fix: cd mcp-unified-consolidated && npm install openai@latest
  2. Unified Server Tools - Most tools return placeholder/mock data

    • Need real implementations for executive, infrastructure, and sync domains
  3. Neon Auth - OAuth flow may timeout

    • Run: neon auth login before using
  4. Hardcoded Paths - Evidence server uses /Users/nb/Evidence-Intake

    • Set EVIDENCE_BASE_PATH in .env to override
  5. Missing .env - No environment file by default

    • Copy .env.example and fill in credentials

Troubleshooting

See TROUBLESHOOTING.md for:

  • Diagnostic script usage
  • Common error fixes
  • LaunchD service issues (macOS)
  • Network timeout solutions
  • Package dependency fixes

Quick diagnostics:

# Run health check
bash diagnostics.sh

# Auto-fix common issues
bash mcp-repair.sh

Development

# Evidence server (watch mode)
cd mcp-evidence-server
npm run dev

# Unified server (watch mode)
cd mcp-unified-consolidated
npm run dev

# MCP Exec (TypeScript compilation)
cd services/mcp-exec
npm run build
npm run dev

Deployment

Cloudflare Workers

Unified server can be deployed as a Cloudflare Worker:

# Deploy to production
wrangler deploy

# Monitor logs
wrangler tail chittymcp

# Set secrets
wrangler secret put CHITTY_ID_TOKEN

See wrangler.toml for configuration.


Evidence Categories

The Evidence Intake server supports 14 categories:

Code Category Description
00 KEY_EXHIBITS High-priority evidence
01 TRO_PROCEEDINGS TRO proceedings
02 LLC_FORMATION Corporate documents
03 MEMBERSHIP_REMOVAL Membership proceedings
04 PREMARITAL_FUNDING Pre-marital property
05 PROPERTY_TRANSACTIONS Real estate
06 FINANCIAL_STATEMENTS Financial docs
07 COURT_FILINGS Court pleadings
08 ATTORNEY_CORRESPONDENCE Attorney letters
09 PERJURY_EVIDENCE Perjury evidence
10 SANCTIONS_RULE137 Sanctions docs
11 COLOMBIAN_PROPERTY Colombian property
12 LEASE_AGREEMENTS Leases
98 DUPLICATES Duplicate files
99 UNSORTED Uncategorized

Files are automatically categorized based on filename patterns.


ChittyID Compliance

CRITICAL: All ChittyIDs must be generated via id.chitty.cc API.

  • Format: CHITTY-{ENTITY}-{SEQUENCE}-{CHECKSUM}
  • Entities: PEO, PLACE, PROP, EVNT, AUTH, INFO, FACT, CONTEXT, ACTOR
  • Never generate IDs locally

Documentation


Architecture

Modular Design

chittymcp/
├── src/
│   ├── core/                    # Shared core utilities
│   │   ├── mcp-server.js       # BaseMCPServer class
│   │   ├── tool-loader.js      # Dynamic tool loading
│   │   ├── chain-executor.js   # Multi-tool workflows
│   │   └── logger.js           # Centralized logging
│   ├── integration/            # ChittyOS service clients
│   │   ├── chittyid-client.js  # ChittyID service
│   │   ├── cloudflare-client.js # Cloudflare API
│   │   └── notion-client.js    # Notion integration
│   ├── tools/                  # Tool modules by domain
│   │   ├── evidence/           # 4 evidence tools
│   │   ├── legal/              # 4 legal tools
│   │   ├── infrastructure/     # 4 infrastructure tools
│   │   └── sync/               # 3 sync tools
│   └── servers/                # Server implementations
│       ├── unified-server.js   # Full consolidated server
│       └── evidence-server.js  # Standalone evidence server
├── config/
│   ├── chains.json             # Workflow chain definitions
│   └── server-config.json      # Server configuration
└── Legacy (backwards compatible):
    ├── mcp-evidence-server/
    └── mcp-unified-consolidated/

Benefits

  • Code Reuse: Shared utilities across all servers
  • Easy Extension: Add new tools without touching core
  • Clean Separation: Tools, integrations, and core separated
  • Dynamic Loading: Tools loaded at runtime
  • Chain Execution: Complex workflows from simple tools

Adding New Tools

  1. Create tool module in src/tools/<category>/:
// src/tools/mycategory/index.js
export const tools = [{
  name: "my_tool",
  description: "What my tool does",
  inputSchema: { /* ... */ }
}];

export { handlers } from "./handlers.js";

// src/tools/mycategory/handlers.js
export const handlers = {
  async my_tool(args) {
    return {
      content: [{
        type: "text",
        text: "Result"
      }]
    };
  }
};
  1. Restart server - tools are loaded automatically!

Workflow Chains

Execute multi-tool workflows defined in config/chains.json:

// Execute a chain
await execute_chain({
  chain_name: "legal-workflow",
  parameters: {
    case_type: "civil",
    documents: ["/path/to/filing.pdf"],
    client_id: "CHITTY-PEO-..."
  }
});

Available chains:

  • executive-decision - Strategic decision-making workflow
  • legal-workflow - Complete legal case management
  • infrastructure-deploy - Cloudflare deployment
  • cross-sync - Device synchronization
  • full-orchestration - All tools combined

Roadmap

v3.0.0 - Completed ✅:

  • [x] Modular architecture with core/tools/integration layers
  • [x] Dynamic tool loading system
  • [x] Chain execution framework
  • [x] 15+ tools across 4 domains
  • [x] ChittyID integration
  • [x] Cloudflare API integration
  • [x] Unified and standalone servers

Next (v3.1.0):

  • [ ] Executive tool AI integration (Anthropic/OpenAI)
  • [ ] ChittyLedger PostgreSQL integration
  • [ ] Hot reload for tool modules
  • [ ] Web API gateway

Future (v4.0.0):

  • [ ] Multi-platform sync (ChatGPT, CustomGPT)
  • [ ] Evidence monitoring automation
  • [ ] Advanced chain orchestration (rollback, parallel execution)

License

MIT License - see LICENSE


Support


Last Updated: 2025-10-30 Version: 3.0.0 ChittyOS Framework: v1.0.1 MCP SDK: 0.5.0

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured