Why use TIM-MCP?

TIM-MCP enhances AI-assisted infrastructure development by:

Steering toward best practices

Without TIM-MCP, AI models might generate generic or outdated Terraform code. TIM-MCP guides models toward IBM-validated patterns and current best practices.

Providing contextual guardrails

TIM-MCP helps AI models navigate the complex IBM Cloud ecosystem with structured access to module interfaces, dependencies, and implementation patterns.

Ensuring accuracy

Foundation models may have limited or outdated knowledge of IBM Cloud modules. TIM-MCP provides real-time access to module details, ensuring accurate parameter usage and proper versioning.

Unlocking distributed knowledge

By connecting models to documentation across many repositories, TIM-MCP helps AI leverage the collective expertise embedded in the Terraform IBM Modules ecosystem.

Install from IBM Project Bob Marketplace

1. Open IBM Project Bob Marketplace:

   • Click the marketplace icon in the Bob pane.
   • Navigate to the MCP tab.

2. Find and install TIM-MCP:

   • Search for "Terraform IBM Modules (TIM)".
   • Click Install on the TIM-MCP server card.

3. Choose installation method:

   • Installation Scope: Select Project (current workspace) or Global (all workspaces).
   • Installation Method: Choose UVX or UVX (Pinned Version).
   • Version:
     • For latest version: Leave as "latest".
     • For pinned version (recommended for production): Enter specific version (e.g., "v1.0.0").

4. Configure GitHub token (optional but recommended):

   • In the GitHub Token field, enter your GitHub personal access token.
   • This helps avoid API rate limits (60 requests/hour without token, 5,000 with token).

5. Complete installation:

   • Click Install.
   • The server will be automatically configured with your settings.

Manual configuration

If you prefer manual configuration or need project-specific settings:

Project-level configuration:

1. Create .bob/mcp.json in your project directory:

   mkdir -p .bob
   

2. Add the configuration:

   {
     "mcpServers": {
       "tim-mcp": {
         "command": "uvx",
         "args": [
           "--from",
           "git+https://github.com/terraform-ibm-modules/tim-mcp.git",
           "tim-mcp"
         ],
         "env": {
           "GITHUB_TOKEN": "your_github_token_here"
         }
       }
     }
   }
   

Global configuration:

1. Click the Settings icon (⚙️) in the top navigation of the Bob pane.
2. Scroll to the bottom of the MCP settings view.
3. Click Edit Global MCP to open mcp_settings.json.
4. Add the TIM-MCP configuration by using the same JSON format as mentioned in the Install TIM-MCP for Claude Desktop section.
5. Verify the installation:
   • Click the Settings icon (⚙️) in the Bob panel → MCP settings to confirm tim-mcp shows as connected.

Quick install

To install TIM-MCP directly into VS Code, click Install TIM-MCP (requires VS Code 1.99+ with GitHub Copilot).

Manual installation

1. Configure TIM-MCP by using one of these methods:

   • Create .vscode/mcp.json in your project directory.
   • Use Command Palette (Ctrl+Shift+P or Cmd+Shift+P) → MCP: Add Server.

2. Add the configuration:

   {
     "servers": {
       "tim-mcp": {
         "command": "uvx",
         "args": [
           "--from",
           "git+https://github.com/terraform-ibm-modules/tim-mcp.git",
           "tim-mcp"
         ],
         "env": {
           "GITHUB_TOKEN": "your_github_token_here"
         }
       }
     }
   }
   

3. Restart VS Code.

4. To start the server, open the Command Palette (Ctrl+Shift+P on Windows or Cmd+Shift+P on macOS), select MCP: Show Servers, right-click tim-mcp, and select Start Server.

Getting started examples

Try these prompts to get started:

Simple virtual server:

I want to create a simple basic virtual server on IBM Cloud and SSH to it

OpenShift cluster:

I am new to IBM Cloud. Help me create a simple and cheap OpenShift cluster and access the console

Enterprise infrastructure examples

For more complex scenarios:

VPC with OpenShift:

Design a VPC + OpenShift: Create a complete container platform with networking,
including multi-zone VPC, subnets, OpenShift/ROKS cluster, and load balancers

Secure landing zone:

Design a Secure Landing Zone: Implement enterprise-grade security with network
isolation, encryption key management, private endpoints, and security groups

Multi-zone HA database:

Design a Multi-Zone HA Database: Design resilient database infrastructure across
3+ availability zones with automated failover, backup strategies, and disaster recovery

Financial Services validated architecture:

Design a FS-Validated Architecture: Deploy compliant infrastructure meeting
Financial Services requirements with HPCS encryption, audit logging, and regulatory controls

Best practices for AI-assisted development

When using TIM-MCP with AI assistants:

1. Start with clear requirements - Describe your infrastructure needs, constraints, and compliance requirements.
2. Review generated code - Always review AI-generated Terraform code before applying.
3. Test in non-production - Deploy to development or staging environments first.
4. Validate configurations - Run terraform plan to review changes before applying.
5. Use version pinning - Pin module versions in production deployments.
6. Follow security best practices - Review security group rules, IAM policies, and encryption settings.
7. Document customizations - Add comments explaining any modifications to generated code.