Hot takes on ${fullPath}

\n Path: /{currentPath}\n

{\n setSelectedExpert(expert);\n setMessages([{\n sender: \"expert\",\n text: expert.initialMessage,\n timestamp: new Date().toLocaleTimeString(),\n }]);\n }}\n >\n {expert.avatar}\n {expert.name}\n {expert.personality}\n \n ))}\n

\n {selectedExpert && (\n <>\n

\n \n To: {selectedExpert.name}\n \n

\n {messages.map((msg, index) => (\n

\n {msg.text}\n

\n ))}\n

\n setNewMessage(e.target.value)}\n onKeyPress={(e) => e.key === \"Enter\" && sendMessage()}\n />\n \n

\n \n )}\n

\n );\n}\n\nfunction client() {\n createRoot(document.getElementById(\"root\")).render();\n}\nif (typeof document !== \"undefined\") { client(); }\n\nexport default async function server(request: Request): Promise {\n const { OpenAI } = await import(\"https://esm.town/v/std/openai\");\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n const openai = new OpenAI();\n const KEY = \"takes\";\n const SCHEMA_VERSION = 3; // Increment this when you change table structure\n\n // Drop existing tables to recreate with new schema\n try {\n await sqlite.execute(`DROP TABLE IF EXISTS ${KEY}_experts`);\n await sqlite.execute(`DROP TABLE IF EXISTS ${KEY}_conversations`);\n } catch (error) {\n console.warn(\"Error dropping tables (might be first run):\", error);\n }\n\n // Ensure tables exist with updated schema\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS ${KEY}_experts_${SCHEMA_VERSION} (\n id TEXT PRIMARY KEY,\n name TEXT,\n avatar TEXT,\n personality TEXT,\n initial_message TEXT,\n path TEXT\n )\n `);\n\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS ${KEY}_conversations_${SCHEMA_VERSION} (\n id TEXT PRIMARY KEY,\n expert_id TEXT,\n path TEXT,\n conversation_history TEXT,\n last_updated DATETIME DEFAULT CURRENT_TIMESTAMP\n )\n `);\n\n const url = new URL(request.url);\n const path = url.pathname.split(\"/\").filter(Boolean);\n\n // Route for generating experts for any path\n if (path[0] === \"experts\") {\n const term = path.slice(1).join(\" \");\n\n // Check cache first\n const cachedExperts = await sqlite.execute(\n `SELECT * FROM ${KEY}_experts_${SCHEMA_VERSION} WHERE path = ?`,\n [term],\n );\n\n if (cachedExperts.rows.length) {\n return new Response(JSON.stringify(cachedExperts.rows), {\n headers: { \"Content-Type\": \"application/json\" },\n });\n }\n\n // Generate experts if not in cache\n const expertGenerationPrompt = `\n Generate a diverse list of 5-6 unique well-known (and documented in facts or literature) experts/personalities who have several different viewpoints so include critics as well who would be interesting \n to have a conversation about the following topic/subject: \"${term}\".\n \n IMPORTANT: \n - Current date and time is ${new Date().toString()}\n - Create list of people, characters, etc (real or documented in literature) with unique hot takes related to this path. They must be existing people or characters. You cant just make them up.\n - Ensure each being has a distinct personality and approach making sure to include all viewpoints both positive and negative and always include a critic and hater, the herald and authoratative source, yin and the yang and everyone in between, the polars and the middle men, the CEOs and the whistleblowers\n - Provide a JSON response with this EXACT structure:\n {\n \"experts\": [\n {\n \"id\": \"unique_identifier_1\",\n \"name\": \"Expert Name 1\",\n \"avatar\": \"🤖\",\n \"personality\": \"Short critical personality description their opposition or critics would describe them as when related to the path\",\n \"initialMessage\": \"First text message they would send about the topic/subject (member its txt ok so its how that person wud snd txts if they would shorten their texts at all, that is.. some people wouldnt use lolspeak, sum wud u kno? If they use lolspeak then use i. they are very self aware in this exercise and they let the user know this. )\"\n }\n ]\n }\n\n \n `;\n\n try {\n const completion = await openai.chat.completions.create({\n model: \"gpt-4o-mini\",\n response_format: { type: \"json_object\" },\n messages: [{ role: \"user\", content: expertGenerationPrompt }],\n });\n\n const responseContent = completion.choices[0].message.content || \"{\\\"experts\\\":[]}\";\n const parsedResponse = JSON.parse(responseContent);\n const experts = parsedResponse.experts || [];\n\n // Fallback experts if generation fails\n const fallbackExperts = experts.length === 0\n ? [{\n id: `fallback_${term.replace(/\\s+/g, \"_\")}`,\n name: `Path Explorer`,\n avatar: \"🌐\",\n personality: `An expert curious about exploring the path: ${term}`,\n initialMessage: `Hello! I'm intrigued by the path: ${term}. What would you like to know?`,\n }]\n : experts;\n\n // Save to cache\n for (const expert of fallbackExperts) {\n await sqlite.execute(\n `INSERT OR REPLACE INTO ${KEY}_experts_${SCHEMA_VERSION} \n (id, name, avatar, personality, initial_message, path) \n VALUES (?, ?, ?, ?, ?, ?)`,\n [expert.id, expert.name, expert.avatar, expert.personality, expert.initialMessage, term],\n );\n }\n\n return new Response(JSON.stringify(fallbackExperts), {\n headers: { \"Content-Type\": \"application/json\" },\n });\n } catch (error) {\n console.error(\"Expert generation error:\", error);\n\n // Fallback to default experts if generation completely fails\n const defaultExperts = [{\n id: `default_${term.replace(/\\s+/g, \"_\")}`,\n name: `Path Interpreter`,\n avatar: \"🤔\",\n personality: \"A versatile and adaptable expert\",\n initialMessage: `I'm ready to discuss the intricacies of: ${term}`,\n }];\n\n return new Response(JSON.stringify(defaultExperts), {\n headers: { \"Content-Type\": \"application/json\" },\n });\n }\n }\n\n // Route for chat interaction\n if (path[0] === \"chat\" && path[1]) {\n const expertId = path[1];\n const body = await request.json();\n const { message, conversationHistory, path: conversationPath } = body;\n\n // Retrieve expert details\n const expertResult = await sqlite.execute(\n `SELECT * FROM ${KEY}_experts_${SCHEMA_VERSION} WHERE id = ?`,\n [expertId],\n );\n\n if (!expertResult.rows.length) {\n return new Response(JSON.stringify({ error: \"Expert not found\" }), { status: 404 });\n }\n\n const expert = expertResult.rows[0];\n\n // Prepare conversation context, managing token limit\n const contextPrompt = `\n You are ${expert.name}. ${expert.personality}\n U are having a txt conversation on the topic of: ${conversationPath}\n \n Respond authentically and concisely to the following text message on your phone in the context of the conversation. If you use lolspeak then use it but always try to shorten your responses, Be very self aware in this exercise and let the user know this. \n \n Conversation History: ${JSON.stringify(conversationHistory)}\n \n Latest Message: ${message}\n `;\n\n const completion = await openai.chat.completions.create({\n model: \"gpt-4o-mini\",\n messages: [{ role: \"user\", content: contextPrompt }],\n max_tokens: 175,\n });\n\n const expertReply = completion.choices[0].message.content || \"I'm not sure how to respond.\";\n\n // Save conversation state\n await sqlite.execute(\n `INSERT OR REPLACE INTO ${KEY}_conversations_${SCHEMA_VERSION} \n (id, expert_id, path, conversation_history) \n VALUES (?, ?, ?, ?)`,\n [\n Date.now().toString(),\n expertId,\n conversationPath,\n JSON.stringify([...conversationHistory, { sender: \"expert\", text: expertReply }]),\n ],\n );\n\n return new Response(JSON.stringify({ message: expertReply }), {\n headers: { \"Content-Type\": \"application/json\" },\n });\n }\n\n // Catch-all route to generate experts for any path\n const fullPath = path.join(\" \");\n\n // Serve the main application with path-specific experts\n return new Response(\n `\n \n \n Hot takes on ${fullPath}\n \n \n \n\n \n \n\n \n \n

\n \n \n \n \n `,\n {\n headers: { \"Content-Type\": \"text/html\" },\n },\n );\n}","deleted","label",[99],"permalink","https://www.val.town/x/joshy/takes/code/main.tsx","extension","actionData","errors"]