Gameball MCP Server
Unified MCP server for Gameball's documentation and API integration, enabling documentation search, code generation, and API operations for customers, transactions, orders, coupons, and more.
README
Gameball Integrations
Unified Model Context Protocol (MCP) server providing comprehensive access to Gameball's developer documentation and API integration.
Features
This server combines two subsystems and a status tool:
Documentation Tools (7 tools)
- No API keys required - Automatically fetches and indexes documentation from GitHub
- Search and browse Gameball developer docs
- Get API endpoint specifications
- Generate code examples in multiple languages (JavaScript, Python, C#, Go, PHP, Java)
- Access tutorials and guides
Integration Tools (71 tools)
- Requires API keys - Make real API calls to Gameball's services
- Customer management (create, update, query, balance, tags, notifications)
- Transaction handling (cashback, redemption, holds, refunds, OTP)
- Order tracking and queries
- Coupon validation and burning
- Batch operations
- Configuration queries
- Referral system
- Leaderboards
- Events and payments
Status Tool (1 tool)
gameball-status- Shows server health, subsystem status, configuration, and cache info
Total: 79 tools in one server
Installation
Prerequisites
- Node.js >= 20
- Clone and build the project:
git clone <repository-url> gameball-integrations
cd gameball-integrations
npm install
npm run build
Add to Your Project (recommended)
Create a .mcp.json file at the root of your project. Claude Code automatically detects this file and loads the MCP server for that workspace.
Docs only (no API keys needed):
{
"mcpServers": {
"gameball": {
"command": "node",
"args": ["/path/to/gameball-integrations/dist/index.js"]
}
}
}
Docs + Integration (requires API keys):
{
"mcpServers": {
"gameball": {
"command": "node",
"args": ["/path/to/gameball-integrations/dist/index.js"],
"env": {
"GAMEBALL_API_KEY": "your-api-key",
"GAMEBALL_SECRET_KEY": "your-secret-key"
}
}
}
}
Tip: If the MCP repo is a sibling directory, use a relative path:
"args": ["../gameball-integrations/dist/index.js"]
Add Globally (all projects)
To make the tools available across all projects, add the config to ~/.claude/settings.json under the same mcpServers structure shown above.
Via Claude Code CLI
# Docs only
claude mcp add gameball node /path/to/gameball-integrations/dist/index.js
# With API keys
claude mcp add gameball node /path/to/gameball-integrations/dist/index.js \
-e GAMEBALL_API_KEY=your-api-key \
-e GAMEBALL_SECRET_KEY=your-secret-key
Environment Variables
| Variable | Required | Description | Default |
|---|---|---|---|
GAMEBALL_API_KEY |
For integration tools | Your Gameball API key | "" |
GAMEBALL_SECRET_KEY |
For sensitive operations | Your Gameball secret key | undefined |
GAMEBALL_BASE_URL |
No | API base URL | https://api.gameball.co |
GITHUB_TOKEN |
No | GitHub token for higher rate limits | undefined |
Tool Categories
Documentation Tools (docs-* prefix)
All documentation tools work without API keys:
| Tool | Description |
|---|---|
docs-lookup-topic |
START HERE - Comprehensive lookup returning tutorials, docs, and API endpoints for a topic |
docs-get-doc |
Get full content of a specific documentation page |
docs-get-api-endpoint |
Get detailed API endpoint specification with parameters and schemas |
docs-generate-code-example |
Generate code examples in JavaScript, Python, C#, Go, PHP, or Java |
docs-search-docs |
Full-text search across all documentation |
docs-list-sections |
Browse documentation structure and available topics |
docs-get-tutorial |
Get step-by-step tutorial content |
Integration Tools (by domain)
All integration tools require GAMEBALL_API_KEY. Tools marked with š also require GAMEBALL_SECRET_KEY.
Authentication (1 tool)
generate-session-token- Generate JWT session token
Customers (21 tools)
create-customer,get-customer,update-customer,delete-customerget-customer-balance,adjust-customer-balancešget-customer-referrals,send-customer-notificationget-customer-tags,add-customer-tag,remove-customer-tag- And more...
Transactions (13 tools)
cashback-transaction,redeem-points,redeem-points-otphold-points,cancel-hold,refund-transactionquery-transactions,reverse-transaction- And more...
Orders (4 tools)
track-order,refund-order,query-orders,cancel-order
Coupons (7 tools)
validate-coupon,burn-coupon,lock-coupon,unlock-couponrevoke-burn-coupon,revoke-validation,list-customer-coupons
Batch Operations (9 tools)
batch-adjust-balance,batch-send-events,batch-cashbackbatch-redeem,batch-refund,batch-hold,batch-cancel-hold- And more...
Configuration (10 tools)
get-action-config,get-challenge-config,get-level-configget-tier-config,query-tiers,query-levels- And more...
Other Domains
- Custom (2 tools): Utilities like customer count, cart tracking
- Events (1 tool): Send customer events
- Leaderboard (1 tool): Get leaderboard rankings
- Payments (1 tool): Record payments
- Referrals (1 tool): Validate referral codes
Status Tool
| Tool | Description |
|---|---|
gameball-status |
Shows subsystem health, API key configuration, cache location, and sync status |
How It Works
Startup
- Fetches the latest commit SHA from GitHub (single API call)
- Loads both subsystems in parallel:
- Docs subsystem: Fetches
.mdxfiles anddocs.json, builds MiniSearch index - Integration subsystem: Fetches
openapi.jsonindependently (gzip-cached)
- Docs subsystem: Fetches
- Parses the OpenAPI spec into endpoint definitions (shared by both subsystems)
- Registers all tools and connects via stdio
Progress is reported to stderr during loading:
Checking for updates...
Loading...
Ready ā 79 tools
Documentation Subsystem
- GitHub Fetching: Fetches Gameball docs from GitHub with batched requests
- Commit-SHA Cache: Validates cache with 1 API call, re-fetches only when docs change
- Full-Text Indexing: MiniSearch with fuzzy matching and heading-based chunking
- Stale-Cache Fallback: Uses cached docs if GitHub is unreachable
Integration Subsystem
- Independent OpenAPI Fetch: Fetches only
openapi.json(not the full docs repo) - Gzip Cache: Stores the OpenAPI spec compressed with atomic writes
- Auto-Generated Tools: All 71 tools generated from OpenAPI spec at startup
- Smart Adaptive Verbosity: Intelligent descriptions save 78% tokens (~7,000 tokens)
- HTTP Client: Wraps Gameball API v4.0 with automatic auth header injection
Graceful Failure
All GitHub fetches have a 10-second timeout. If GitHub is unavailable:
- With cache: Falls back to stale cached data
- Without cache (cold start): Server starts with 0 tools + status tool
- The
gameball-statustool always registers and shows the error state
First run: ~2-3 seconds (fetches from GitHub) Subsequent runs: ~200-500ms (validates cache with 1 API call)
Architecture
gameball-integrations/
āāā src/
ā āāā index.ts # Entry point (stdio transport)
ā āāā server.ts # Server creation (parallel init, graceful failure)
ā āāā status-tool.ts # gameball-status tool
ā āāā docs/ # Documentation subsystem (7 tools)
ā ā āāā index.ts # Init + tool registration
ā ā āāā fetcher.ts # GitHub fetching + SHA-based cache
ā ā āāā doc-indexer.ts # MiniSearch indexing + chunking
ā ā āāā mdx-parser.ts # MDX ā plain text
ā ā āāā codegen.ts # Multi-language code generation
ā ā āāā types.ts # Doc-specific types
ā ā āāā tools/ # 7 documentation tools
ā āāā integration/ # API integration subsystem (71 tools)
ā ā āāā index.ts # Client creation + tool registration
ā ā āāā openapi-fetcher.ts # Independent openapi.json fetch + gzip cache
ā ā āāā http-client.ts # GameballClient (auth, error formatting)
ā ā āāā generator.ts # Auto-generates 71 tools from OpenAPI
ā āāā shared/ # Common utilities
ā āāā github-cache.ts # GitHub fetch, SHA check, cache dir, timeout
ā āāā types.ts # Shared type definitions (ApiEndpoint, etc.)
ā āāā openapi-parser.ts # OpenAPI JSON ā ApiEndpoint map
Cache Layout
~/.cache/gameball-integrations/
āāā manifest.json # Docs cache manifest (commitSha, file list)
āāā [__-delimited .mdx files] # Cached documentation pages
āāā openapi-manifest.json # OpenAPI cache manifest (commitSha, timestamp)
āāā openapi.json.gz # Gzip-compressed OpenAPI spec
Usage Examples
Learning about Gameball
Use docs-lookup-topic to learn about "customer referrals"
Use docs-get-doc to read "tutorials-new/referral-setup"
Use docs-generate-code-example for JavaScript customer creation
Making API Calls
Use create-customer to register a new customer
Use cashback-transaction to reward points
Use query-transactions to check transaction history
Checking Server Health
Use gameball-status to see subsystem status, config, and cache info
Development
# Install dependencies
npm install
# Run in development mode (with auto-reload)
npm run dev
# Build for production
npm run build
# Run built version
npm start
Requirements
- Node.js >= 20
- For integration tools: Valid Gameball API credentials
License
MIT
Links
Support
For issues or questions:
- Documentation tools: Check cache at
~/.cache/gameball-integrations/ - Integration tools: Verify API credentials in environment variables
- GitHub Issues: [Report bugs or request features]
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.