Skip to main content
Glama

run_ai_action

Execute AI prompts on specific Carbon Voice messages to generate responses, analyze content, or perform automated actions based on message IDs.

Instructions

Run an AI Action (Prompt) for a message. You can run an AI Action for a message by its ID or a list of message IDs.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
prompt_idYes
message_idsYes
channel_idNo
workspace_idNo
languageNoThe language of the response. Defaults to original message.

Implementation Reference

  • The handler function that implements the core logic of the 'run_ai_action' tool. It invokes the simplified API's aIResponseControllerCreateResponse method with the input arguments and authentication header, formats the response, and handles errors.
    async (args: CreateAIResponse, { authInfo }): Promise<McpToolResponse> => { try { return formatToMCPToolResponse( await simplifiedApi.aIResponseControllerCreateResponse( args, setCarbonVoiceAuthHeader(authInfo?.token), ), ); } catch (error) { logger.error('Error running ai action:', { error }); return formatToMCPToolResponse(error); } },
  • src/server.ts:861-885 (registration)
    Registers the 'run_ai_action' tool on the MCP server, specifying its name, description, input schema, annotations, and handler function.
    server.registerTool( 'run_ai_action', { description: 'Run an AI Action (Prompt) for a message. You can run an AI Action for a message by its ID or a list of message IDs.', inputSchema: aIResponseControllerCreateResponseBody.shape, annotations: { readOnlyHint: false, destructiveHint: false, }, }, async (args: CreateAIResponse, { authInfo }): Promise<McpToolResponse> => { try { return formatToMCPToolResponse( await simplifiedApi.aIResponseControllerCreateResponse( args, setCarbonVoiceAuthHeader(authInfo?.token), ), ); } catch (error) { logger.error('Error running ai action:', { error }); return formatToMCPToolResponse(error); } }, );
  • Zod schema definition for the input body of the aIResponseControllerCreateResponse API call, used as inputSchema for the tool.
    export const aIResponseControllerCreateResponseBody = zod.object({ "prompt_id": zod.string(), "message_ids": zod.array(zod.string()), "channel_id": zod.string().optional(), "workspace_id": zod.string().optional(), "language": zod.string().optional().describe('The language of the response. Defaults to original message.') })
  • TypeScript interface defining the shape of the input arguments for the 'run_ai_action' handler.
    export interface CreateAIResponse { prompt_id: string; message_ids: string[]; channel_id?: string; workspace_id?: string; /** The language of the response. Defaults to original message. */ language?: string; }

Latest Blog Posts

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/PhononX/cv-mcp-server'

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