Skip to main content
Glama

MCP Prompts Server

05-templates-guide.md2.24 kB
# Templates Guide Templates are reusable prompt patterns with variables, enabling dynamic content generation for LLM workflows. In MCP-Prompts, templates are first-class resources, accessible via API, CLI, and workflows. > **MCP-Prompts** uses hexagonal architecture: template logic is part of the core domain, with adapters for storage and transport. See [Overview](00-overview.md) and [API Reference](04-api-reference.md) for details. ## Variable Syntax You can use variables in your prompt templates using double curly braces: ```text Hello, {{name}}! Today is {{date}}. ``` ## Extracting Variables Programmatically The server can extract variables from prompt content and require them at runtime. Example: ```json { "name": "Greeting Template", "content": "Hello, {{name}}! Today is {{date}}.", "isTemplate": true, "variables": ["name", "date"] } ``` ## Applying a Template Suppose you have a template as above. To apply it: ```json { "templateId": "greeting-template", "variables": { "name": "Alice", "date": "2024-06-19" } } ``` Result: ``` Hello, Alice! Today is 2024-06-19. ``` ## Advanced: Multi-Step Workflow Example You can chain prompts by referencing outputs as variables in subsequent steps. Example: 1. **Step 1: Generate summary** ```json { "id": "summarize", "content": "Summarize the following text: {{input}}", "isTemplate": true, "variables": ["input"] } ``` 2. **Step 2: Generate questions** ```json { "id": "questions", "content": "Based on this summary: {{summary}}, generate 3 questions.", "isTemplate": true, "variables": ["summary"] } ``` **Workflow:** - Apply `summarize` with your input text. - Take the output, use as `summary` variable for `questions`. ## Embedding Resources via MCP URIs You can reference external resources in your prompt content: ```text Use the following context: [resource](mcp://resource-id) ``` ## Exporting to Multiple Formats Prompts and templates can be exported as: - MDC (Cursor Rules) - PGAI (Postgres/JSON) - Plain text See the CLI or API for export commands. --- > For more advanced workflow and chaining, see the [examples/](../examples/) directory and integration guides.

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/sparesparrow/mcp-prompts'

If you have feedback or need assistance with the MCP directory API, please join our Discord server