Resources
Resources are read-only content that AI clients can access via MCP. Create reference materials, documentation snippets, or any context that your AI assistant can retrieve when needed.
Overview
Resources provide a way to store static content that the AI can read but not modify. Unlike commands (which are prompts) or skills (which are instructions), resources are reference materials the AI can consult.
Common use cases include:
- API documentation — Store API specs or endpoint references
- Code snippets — Reference implementations or patterns
- Configuration templates — Standard config files or schemas
- Style guides — Writing or coding style documentation
- Project context — Architecture decisions or project-specific info
Creating a Resource
To create a new resource:
- Navigate to your workspace
- Click the Resources tab
- Click New Resource
- Fill in the resource details:
- Name — The resource name in kebab-case (e.g.,
api-reference) - Description — A brief description of what this resource contains
- Content — The actual content of the resource
- Name — The resource name in kebab-case (e.g.,
- Click Create
Accessing Resources
Resources are exposed via MCP as read-only content. When connected to your AI IDE, the AI can access resources by name to retrieve their content.
Resources use the MCP resources protocol, allowing AI clients to list available resources and read their contents. This is different from commands (prompts) and skills (instructions).
Enabling & Disabling Resources
Resources can be enabled or disabled:
- Open the resource card menu (three dots)
- Click Enable or Disable
- Disabled resources appear dimmed and won't be accessible via MCP
Exporting Resources
Resources can be exported as JSON files for backup or sharing:
- Open the resource card menu (three dots)
- Click Export as JSON
- The resource will be downloaded as a JSON file
The exported JSON format:
{
"name": "api-reference",
"description": "REST API endpoint documentation",
"content": "## Users API\n\nGET /users - List all users\nPOST /users - Create a user\n..."
}
Importing Resources
Import resources from JSON files:
- Click the Import button next to "New Resource"
- Drag and drop a JSON file or click to browse
- The system validates the JSON structure
- Preview shows resource name and description
- Click Import Resource to create it
Note: If a resource with the same name exists, it will be auto-renamed (e.g., "api-reference" → "api-reference-2").
Content Limits
Resources have the following content limits:
| Field | Limit |
|---|---|
| Name | 30 characters |
| Description | 100 characters |
| Content | 10,000 characters |
Workspace Limits
Each workspace can have up to 200 resources. If you reach this limit, you'll need to delete unused resources before creating new ones.
Resources vs Commands
Understanding when to use resources versus commands:
| Feature | Resources | Commands |
|---|---|---|
| Purpose | Reference content the AI reads | Prompts that trigger AI actions |
| MCP Type | Resources (read-only) | Prompts (executable) |
| Arguments | No | Yes (up to 3) |
| Use case | Documentation, context | Actions, workflows |