Paprika MCP Server

Instructions

Implementation Reference

• src/paprika_mcp/tools/read_recipe.py:10-110 (handler)

  The handler function that executes the read_recipe tool: retrieves recipe by ID or title, formats all fields into markdown text.
  async def read_recipe_tool(args: dict[str, Any]) -> list[TextContent]: """Read full recipe data by ID or exact title.""" recipe_id = args.get("id") recipe_title = args.get("title") if not recipe_id and not recipe_title: return [ TextContent(type="text", text="Error: Must provide either 'id' or 'title'") ] # Get the remote remote = get_remote() # If we have an ID, use it directly if recipe_id: # Need to get all recipes first to find the hash for r in remote.recipes: if r.uid == recipe_id: recipe = r break if not recipe: return [ TextContent( type="text", text=f"Error: No recipe found with ID '{recipe_id}'", ) ] else: # Search by title # NOTE: This is not optimal - ideally we'd cache the recipe list # and search through it, but for now we fetch all recipes all_recipes = remote.recipes # Normalize the search title normalized_search = normalize_string(recipe_title) # Find matching recipe recipe = None for r in all_recipes: if normalize_string(r.name) == normalized_search: recipe = r break if not recipe: return [ TextContent( type="text", text=f"Error: No recipe found with title '{recipe_title}'", ) ] # Format the recipe data output_lines = [ f"# {recipe.name}", "", f"**UID:** {recipe.uid}", ] if recipe.description: output_lines.extend(["", "## Description", recipe.description]) if recipe.categories: output_lines.extend(["", "## Categories", ", ".join(recipe.categories)]) if recipe.source: output_lines.extend(["", "## Source", recipe.source]) if recipe.source_url: output_lines.extend(["", "## Source URL", recipe.source_url]) if recipe.prep_time: output_lines.extend(["", f"**Prep Time:** {recipe.prep_time}"]) if recipe.cook_time: output_lines.extend(["", f"**Cook Time:** {recipe.cook_time}"]) if recipe.total_time: output_lines.extend(["", f"**Total Time:** {recipe.total_time}"]) if recipe.servings: output_lines.extend(["", f"**Servings:** {recipe.servings}"]) if recipe.difficulty: output_lines.extend(["", f"**Difficulty:** {recipe.difficulty}"]) if recipe.rating: output_lines.extend(["", f"**Rating:** {recipe.rating}"]) if recipe.ingredients: output_lines.extend(["", "## Ingredients", recipe.ingredients]) if recipe.directions: output_lines.extend(["", "## Directions", recipe.directions]) if recipe.notes: output_lines.extend(["", "## Notes", recipe.notes]) if recipe.nutritional_info: output_lines.extend(["", "## Nutritional Info", recipe.nutritional_info]) return [TextContent(type="text", text="\n".join(output_lines))]
• src/paprika_mcp/tools/read_recipe.py:113-131 (schema)

  The TOOL_DEFINITION dictionary defining the tool name, description, and input schema for read_recipe.
  # Tool definition TOOL_DEFINITION = { "name": "read_recipe", "description": ( "Read full recipe data by ID or exact title. " "Returns all recipe fields including categories, times, ingredients, directions, etc." ), "inputSchema": { "type": "object", "properties": { "id": {"type": "string", "description": "Recipe UID to read"}, "title": { "type": "string", "description": "Exact recipe title to read (alternative to id)", }, }, "required": [], }, }
• src/paprika_mcp/tools/__init__.py:18-21 (registration)

  Registration of the read_recipe tool in the TOOLS dictionary, linking definition and handler.
  "read_recipe": { "definition": READ_RECIPE_DEF, "handler": read_recipe_tool, },
• src/paprika_mcp/tools/__init__.py:5-6 (registration)

  Imports the tool definition and handler from read_recipe.py for registration.
  from .read_recipe import TOOL_DEFINITION as READ_RECIPE_DEF from .read_recipe import read_recipe_tool