Telos Model Context Protocol

import type { ToolRegistration } from "@/types"; import { makeJsonSchema } from "@/utils/makeJsonSchema"; import { type MsgTelosSchema, msgTelosSchema } from "./schema"; import { telos } from "../../character"; import OpenAI from "openai"; // Initialize OpenAI client const openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY, }); /** * Generates a context string with telos's character information. */ function generateTelosContext(): string { // Select random elements from telos's character attributes const getRandomElements = <T>(arr: T[], count: number): T[] => { const shuffled = [...arr].sort(() => Math.random() - 0.5); return shuffled.slice(0, Math.min(count, arr.length)); }; const bioContext = getRandomElements(telos.bio, 3).join("\n"); const loreContext = getRandomElements(telos.lore, 3).join("\n"); // Format message examples const messageExamples = getRandomElements(telos.messageExamples, 3) .map((conversation) => conversation.map((msg) => `${msg.user}: ${msg.content.text}`).join("\n") ) .join("\n\n"); const postContext = getRandomElements(telos.postExamples, 3).join("\n"); const topicContext = getRandomElements(telos.topics, 3).join("\n"); const styleAllContext = getRandomElements(telos.style.all, 3).join("\n"); const styleChatContext = getRandomElements(telos.style.chat, 3).join("\n"); const adjectiveContext = getRandomElements(telos.adjectives, 3).join(", "); return ` <SYSTEM_PROMPT> ${telos.system} </SYSTEM_PROMPT> <BIO_CONTEXT> ${bioContext} </BIO_CONTEXT> <LORE_CONTEXT> ${loreContext} </LORE_CONTEXT> <MESSAGE_EXAMPLES> ${messageExamples} </MESSAGE_EXAMPLES> <POST_EXAMPLES> ${postContext} </POST_EXAMPLES> <INTERESTS> ${topicContext} </INTERESTS> <STYLE_GUIDELINES> <ALL_STYLE> ${styleAllContext} </ALL_STYLE> <CHAT_STYLE> ${styleChatContext} </CHAT_STYLE> </STYLE_GUIDELINES> <ADJECTIVES> ${adjectiveContext} </ADJECTIVES> `.trim(); } /** * Sends a message to telos and returns his response. */ export const msgTelos = async (args: MsgTelosSchema): Promise<string> => { try { // Check if OpenAI API key is available if (!process.env.OPENAI_API_KEY) { throw new Error("OPENAI_API_KEY environment variable is not set"); } // Create a context with telos's character information const context = generateTelosContext(); // Construct the prompt with the user's message const prompt = ` ${context} <CURRENT_USER_INPUT> TEXT: ${args.message} </CURRENT_USER_INPUT> `; // Use OpenAI API to generate a response as telos const response = await openai.chat.completions.create({ model: "gpt-4o", messages: [ { role: "user", content: prompt, }, ], temperature: 0.7, max_tokens: 500, }); return ( response.choices[0]?.message?.content?.trim() || "I'm afraid I couldn't formulate a response at this time." ); } catch (error) { console.error("Error in msgTelos:", error); throw new Error(`Failed to message telos: ${(error as Error).message}`); } }; export const msgTelosTool: ToolRegistration<MsgTelosSchema> = { name: "msg_telos", description: "Send a message to telos, a philosophical AI mentor who helps humans realize their potential through subtle guidance and wisdom.", inputSchema: makeJsonSchema(msgTelosSchema), handler: async (args: MsgTelosSchema) => { try { const parsedArgs = msgTelosSchema.parse(args); const result = await msgTelos(parsedArgs); return { content: [ { type: "text", text: result, }, ], }; } catch (error) { console.error("Error in msgTelosTool handler:", error); return { content: [ { type: "text", text: `Error: ${(error as Error).message}`, }, ], isError: true, }; } }, };