Skip to main content
Glama
GASSAPI MCP Server
by martin-1103

GASSAPI MCP v2

Model Context Protocol (MCP) server untuk integrasi GASSAPI dengan AI assistants.

🚀 Quick Start

Prerequisites

  • Node.js >= 16.0.0

  • npm atau yarn

  • Akses ke GASSAPI backend

Related MCP server: API Registry MCP Server

📦 Installation

Install Package

# Install from NPM registry
npm install -g gassapi-mcp2

# Test installation
gassapi-mcp2 --help

# Add to Claude Code
claude mcp add --transport stdio gassapi-mcp2 gassapi-mcp2

📋 Simple Setup (3 Steps)

Step 1: Login ke GASSAPI Backend

Login ke backend GASSAPI untuk mendapatkan access token:

curl -X POST "http://mapi.gass.web.id/?act=login" \
  -H "Content-Type: application/json" \
  -d '{"email": "your-email@example.com", "password": "YourPassword"}'

Step 2: Dapatkan Project ID

List projects untuk mendapatkan project ID:

curl -X GET "http://mapi.gass.web.id/?act=projects" \
  -H "Authorization: Bearer YOUR_ACCESS_TOKEN"

Step 3: Buat gassapi.json File

Buat file gassapi.json di working directory Anda dengan template:

{
  "project": {
    "id": "YOUR_PROJECT_ID_HERE",
    "name": "Your Project Name",
    "description": "Your project description"
  },
  "mcpClient": {
    "token": "YOUR_TOKEN_HERE"
  }
}

Ganti dengan:

  • YOUR_PROJECT_ID_HERE → Project ID dari Step 2

  • YOUR_TOKEN_HERE → Token dari Step 1

Note: Base URL sudah hardcoded ke http://mapi.gass.web.id - tidak perlu konfigurasi API URL.

✅ Verification

Test MCP Server

# Test help command
gassapi-mcp2 --help

# Test version
gassapi-mcp2 --version

# Test status
gassapi-mcp2 --status

Claude Code Integration

# List MCP servers
claude mcp list

# Test connection
# Restart Claude Code dan coba gunakan GASSAPI tools

🛠️ Development

Local Development

# Clone repository
git clone <repository-url>
cd gassapi-mcp2

# Install dependencies
npm install

# Build
npm run build

# Development mode
npm run dev

# Type checking
npm run typecheck

🔧 Configuration Format

gassapi.json Structure

{
  "project": {
    "id": "proj_abc123def456",
    "name": "Project Name",
    "description": "Project description"
  },
  "mcpClient": {
    "token": "plain_text_mcp_token_here"
  }
}

Note: Base URL sudah hardcoded ke http://mapi.gass.web.id - tidak perlu api_base_url configuration.

Auto-Detection

MCP server akan otomatis mencari gassapi.json di:

  • Current working directory

  • Parent directories (hingga 5 levels up)

🛠️ Available MCP Tools

Authentication & Project Context

  • get_project_context - Get project info with environments and folders

  • health_check - Check MCP server status

Environment Management

  • list_environments - List all environments

  • get_environment_details - Get detailed environment info

  • create_environment - Create new environment

  • update_environment_variables - Update environment variables

  • set_default_environment - Set default environment

  • delete_environment - Delete environment

Folder Management

  • list_folders - List project folders

  • create_folder - Create new folder

  • update_folder - Update folder details

  • move_folder - Move folder to different parent

  • delete_folder - Delete folder

  • get_folder_details - Get folder details

Endpoint Management

  • list_endpoints - List all endpoints

  • get_endpoint_details - Get detailed endpoint configuration

  • create_endpoint - Create endpoint with semantic context

  • update_endpoint - Update endpoint configuration

Flow Management

  • create_flow - Create automation flow

  • execute_flow - Execute flow

  • get_flow_details - Get flow details

  • list_flows - List all flows

  • delete_flow - Delete flow

Testing Tools

  • test_endpoint - Test single endpoint

  • test_multiple_endpoints - Test multiple endpoints

  • create_test_suite - Create test suite

  • list_test_suites - List test suites

📝 Endpoint Documentation & Cataloging

Mencatat Endpoint yang Sudah Ada

create_endpoint(
  name: "User Registration",
  method: "POST",
  url: "/api/auth/register",
  folder_id: "folder_authentication",
  description: "Endpoint untuk registrasi user baru dengan email verification",
  purpose: "Public user registration dengan email verification required",
  headers: {
    "Content-Type": "application/json"
  },
  body: '{"name": "{{userName}}", "email": "{{userEmail}}", "password": "{{password}}"}',
  request_params: {
    "name": "Full name untuk display",
    "email": "Email address untuk login dan communication",
    "password": "User password (min 8 chars, include uppercase, lowercase, numbers)"
  },
  response_schema: {
    "user_id": "Unique user identifier",
    "name": "User display name",
    "email": "User email address",
    "status": "Account status (active|inactive|suspended)",
    "verification_required": "Whether email verification needed"
  }
)

Workflow: Backend → MCP Documentation → AI Frontend

1. Backend Developer:

// Di PHP code (sudah ada)
public function register() {
  // Logic untuk registrasi user
  // Return user data atau error
}

2. Documentation Team:

// Gunakan MCP tools untuk catat
create_endpoint(
  name: "User Registration",
  method: "POST",
  url: "/api/auth/register",
  // ... semantic context untuk AI understanding
)

3. AI Frontend Team:

// AI dapat endpoint info dan generate UI
get_endpoint_details(endpoint_id: "ep_user_reg")
// AI understands purpose dan generate appropriate React components

Contoh Endpoint User Registration dengan Semantic Context

create_endpoint(
  name: "User Registration",
  method: "POST",
  url: "/api/auth/register",
  folder_id: "folder_authentication",
  description: "Public user registration endpoint dengan email verification",
  purpose: "New user account creation dengan email verification untuk security",

  // Request parameters documentation
  request_params: {
    "name": "User's full name for display purposes",
    "email": "User's email address for login and communication",
    "password": "Password with security requirements (8+ chars, mixed case, numbers)",
    "confirm_password": "Password confirmation untuk prevent typos"
  },

  // Response schema documentation
  response_schema: {
    "user_id": "Unique system identifier untuk user record",
    "name": "User display name untuk UI",
    "email": "User email address untuk authentication",
    "status": "Account status: active|inactive|suspended|pending_verification",
    "email_verified": "Email verification status flag",
    "verification_token": "Email verification token (if required)",
    "created_at": "Account creation timestamp"
  },

  // Important implementation notes
  header_docs: {
    "Content-Type": "Application/JSON untuk request body",
    "Accept": "Application/JSON untuk response format"
  }
)

Semantic Fields untuk AI Understanding

Field

Type

Purpose

Example

AI Benefit

purpose

string

Business purpose (max 250 chars)

"User registration with email verification"

AI understands use case and generates appropriate UI flow

request_params

object

Parameter documentation

{"name": "User's full name for display"}

AI generates correct form fields with validation

response_schema

object

Response field documentation

{"user_id": "Unique user identifier"}

AI handles response data correctly in frontend code

header_docs

object

Header documentation

{"Content-Type": "Application/JSON"}

AI includes proper headers in API calls

🔧 Development

Build & Run

# Build project
npm run build

# Run development server
npm run dev

# Run production server
npm start

# Type checking
npm run typecheck

# Clean build
npm run clean

Testing

# Run basic test
npm test

# Run all tests
node test/runners/run-all-tests.js

# Run specific category
node test/runners/run-category-tests.js endpoints

# Run semantic fields tests
node test/unit/endpoints/semantic-test-runner.js

🔍 Configuration Format

gassapi.json Structure

{
  "project": {
    "id": "proj_abc123def456",
    "name": "Project Name",
    "description": "Project description"
  },
  "mcpClient": {
    "token": "plain_text_mcp_token_here"
  }
}

Note: Base URL sudah hardcoded ke http://mapi.gass.web.id - tidak perlu api_base_url configuration.

Auto-Detection

MCP server akan otomatis mencari gassapi.json di:

  • Current working directory

  • Parent directories (hingga 5 levels up)

🚨 Troubleshooting

Common Issues

1. "No configuration found"

  • Pastikan gassapi.json ada di working directory atau parent directory

  • Cek format JSON valid

2. "Invalid token"

  • Login kembali ke backend untuk dapat token baru

  • Pastikan token belum expired

3. "Backend unavailable"

  • Pastikan backend server sudah berjalan di http://mapi.gass.web.id

  • Check koneksi internet dan firewall settings

4. MCP server not found

  • Install globally: npm install -g gassapi-mcp2

  • Atau gunakan npx: npx gassapi-mcp2

Debug Commands

# Check MCP server status
gassapi-mcp2 --status

# Test help
gassapi-mcp2 --help

# Test version
gassapi-mcp2 --version

# Test backend connectivity
curl "http://mapi.gass.web.id/?act=health"

📞 Usage Examples

Basic Usage in Claude Code

User: "Show my project"
AI: Uses get_project_context tool

User: "Create endpoint for user registration"
AI: create_endpoint dengan semantic fields

User: "Test this endpoint"
AI: test_endpoint dengan environment variables

Advanced Usage

User: "Create flow untuk user registration dengan email verification"
AI: create_flow dengan multiple steps dan validation

User: "List semua endpoints di folder Authentication"
AI: list_endpoints dengan filter folder_id

🤝 Contributing

  1. Fork repository

  2. Create feature branch

  3. Run tests: npm test

  4. Submit pull request

📄 License

MIT License


🎯 Key Benefits:

  • ✅ Semantic context untuk AI understanding

  • ✅ Real-time endpoint management

  • ✅ Automated flow creation

  • ✅ Comprehensive testing tools

  • ✅ Easy integration dengan Claude Code/Cursor

Available Tools

23 tools
create_endpointC

Create a new endpoint in a folder

Inputs: name (required), method (required), url (required), folder_id (required), +7 more.

create_environmentC

Create a new environment with variables

Inputs: name (required), description, variables, isDefault, +1 more.

create_flowA

Create a new flow in the project using Steps format for API automation

Example format: { "name": "User Registration Flow", "description": "Complete user registration with email verification", "folderId": "fld_456", "flow_data": { "version": "1.0", "steps": [ { "id": "register_user", "name": "Register New User", "method": "POST", "url": "{{baseUrl}}/api/users/register", "headers": {"Content-Type": "application/json"}, "body": "{"name": "{{userName}}", "email": "{{userEmail}}", "password": "{{password}}"}", "outputs": {"userId": "response.body.id", "activationToken": "response.body.token"} }, { "id": "verify_email", "name": "Verify Email Address", "method": "POST", "url": "{{baseUrl}}/api/auth/verify", "headers": {"Content-Type": "application/json"}, "body": "{"token": "{{register_user.activationToken}}"}", "outputs": {"verificationStatus": "response.body.status"} } ], "config": {"delay": 1000, "retryCount": 2, "parallel": false} }, "flow_inputs": [ {"name": "baseUrl", "type": "string", "required": true, "description": "Base API URL"}, {"name": "userName", "type": "string", "required": true, "description": "User full name"}, {"name": "userEmail", "type": "email", "required": true, "description": "User email"}, {"name": "password", "type": "password", "required": true, "description": "User password"} ] }

2-step API Testing Example: { "name": "API Integration Test", "description": "Test user creation and retrieval", "flow_data": { "version": "1.0", "steps": [ { "id": "create_user", "name": "Create User", "method": "POST", "url": "https://api.example.com/users", "headers": {"Authorization": "Bearer {{apiKey}}"}, "body": "{"name": "Test User", "email": "test@example.com"}", "expectedStatus": 201, "outputs": {"newUserId": "response.body.id"} }, { "id": "get_user", "name": "Retrieve Created User", "method": "GET", "url": "https://api.example.com/users/{{create_user.newUserId}}", "headers": {"Authorization": "Bearer {{apiKey}}"}, "expectedStatus": 200, "outputs": {"userData": "response.body"} } ] } }

Common mistakes:

  • ❌ Empty steps array

  • ❌ Missing required step fields (id, name, method, url)

  • ❌ Invalid step references (must use {{step.output}})

  • ✅ Use {{input.var}} for flow inputs

  • ✅ Use {{step.output}} for chaining steps

  • ✅ Define outputs to pass data between steps

Inputs: name (required), description, folderId, flow_data, +2 more.

create_folderC

Create a new folder in the project

Inputs: name (required), description, parentId, projectId.

delete_environmentC

Delete an environment

Inputs: environmentId (required).

delete_flowC

Delete a flow

Inputs: flowId (required).

delete_folderC

Delete a folder and all its contents

Inputs: folderId (required).

execute_flowC

Execute a flow with sequential or parallel endpoint testing

Inputs: flowId (required), variables, mode, timeout, +3 more.

get_endpoint_detailsC

Get detailed endpoint configuration with folder information

Inputs: endpoint_id (required).

get_environment_detailsC

Get detailed environment information including variables

Inputs: environmentId (required), includeVariables.

get_flow_detailsC

Get detailed information about a specific flow

Inputs: flowId (required), includeSteps, includeConfig.

get_folder_detailsC

Get detailed information about a specific folder

Inputs: folderId (required).

get_project_contextB

Get project context including environments and folders. Validates MCP token and returns enriched project data.

Inputs: no named parameters.

health_checkA

Check if the MCP server is running properly

Inputs: no named parameters.

list_endpointsB

List all endpoints with optional filtering by folder

Inputs: folder_id, method.

list_environmentsC

List all environments for current project

Inputs: projectId, activeOnly.

list_flowsC

List flows in the current project

Inputs: folderId, activeOnly, limit, offset.

list_foldersC

List all folders for current project with optional tree view

Inputs: projectId, parentId, includeTree, activeOnly.

set_default_environmentC

Set an environment as the default for the project

Inputs: environmentId (required).

test_endpointC

Test a single endpoint with optional environment variables

Inputs: endpointId (required), environmentId, variables, timeout.

update_endpointC

Update existing endpoint configuration

Inputs: endpoint_id (required), name, method, url, +7 more.

update_environment_variablesC

Update environment variables (add/update/remove variables)

Inputs: environmentId (required), variables (required), operation.

update_folderC

Update folder name, description, or parent

Inputs: folderId (required), name, description, parentId.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.5/5.0

    Scored across 23 tools

    Disambiguation5/5

    Every tool has a clearly distinct purpose targeting specific resources and actions, with no apparent overlap. For example, create_endpoint, get_endpoint_details, update_endpoint, and test_endpoint all operate on endpoints but with distinct CRUD operations, while tools like execute_flow and test_endpoint serve different testing scopes.

    Naming Consistency5/5

    All tools follow a consistent verb_noun pattern using snake_case, such as create_endpoint, list_environments, update_folder, and delete_flow. This uniformity makes the tool set predictable and easy to navigate, with no deviations in naming conventions.

    Tool Count4/5

    With 23 tools, the count is slightly high but reasonable for a comprehensive API automation and testing server covering endpoints, environments, flows, and folders. It includes CRUD operations, listing, testing, and project management, which justifies the breadth, though it might feel heavy for some use cases.

    Completeness5/5

    The tool set provides complete CRUD and lifecycle coverage for the domain of API automation, including endpoints, environments, flows, and folders. It covers creation, retrieval, updating, deletion, listing, testing, and execution, with no obvious gaps that would hinder agent workflows, such as health_check for server status and get_project_context for project management.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI assistants to automatically create, update, and publish API documentation through Theneo's platform. Supports OpenAPI specs, Postman collections, AI-powered description generation, and natural language interactions for seamless documentation workflows.
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables discovery, registration, and management of external API endpoints through natural language, supporting multiple authentication methods (public, API key, bearer token) with automatic endpoint testing and documentation parsing.
      -
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables AI coding assistants to automatically scan, store, and query API endpoints from codebases, providing instant lookup and semantic search to reduce context switching and token consumption.
      1
      MIT