Generative Art Scene Interpreter

Provide a high-level creative prompt. An AI architect will design a formal specification for the scene, which will then be rendered.

\n\n

\n \n

\n\n

\n \n

Architect's Justification

Awaiting scene specification...

\n\n\n\n`;\n}\n\n// --- VAL TOWN HTTP HANDLER ---\n\n/**\n * Extracts a JSON object from a string, guarding against\n * conversational text from the AI.\n * @param {string} text - The raw response text from the AI.\n * @returns {string | null} The extracted JSON string, or null.\n */\nfunction extractJson(text: string): string | null {\n // Find the first '{' and the last '}'\n const firstBrace = text.indexOf(\"{\");\n const lastBrace = text.lastIndexOf(\"}\");\n\n if (firstBrace === -1 || lastBrace === -1 || lastBrace < firstBrace) {\n return null;\n }\n // Extract the substring that looks like a JSON object\n const jsonString = text.substring(firstBrace, lastBrace + 1);\n return jsonString;\n}\n\nexport default async function (req: Request) {\n const url = new URL(req.url);\n\n // --- POST Request: Handle AI Generation ---\n if (req.method === \"POST\") {\n try {\n const openai = new OpenAI();\n const body = await req.json();\n const userPrompt = body.prompt;\n\n if (\n !userPrompt || typeof userPrompt !== \"string\" ||\n userPrompt.trim().length === 0\n ) {\n return new Response(\n JSON.stringify({ error: \"A valid 'prompt' is required.\" }),\n {\n status: 400,\n headers: { \"Content-Type\": \"application/json\" },\n },\n );\n }\n\n const completion = await openai.chat.completions.create({\n model: \"gpt-4o\",\n messages: [\n { role: \"system\", content: SCENE_GENERATOR_PROMPT },\n { role: \"user\", content: userPrompt },\n ],\n temperature: 0.7,\n response_format: { type: \"json_object\" },\n });\n\n const aiContent = completion.choices[0].message.content;\n\n // Sanitize AI output: Do not trust the AI to *only* return JSON.\n // We must extract it from potential conversational chatter.\n const jsonString = extractJson(aiContent);\n\n if (!jsonString) {\n console.error(\"AI returned invalid/non-JSON content:\", aiContent);\n throw new Error(\"AI failed to return a valid JSON specification.\");\n }\n\n // Final validation to ensure core fields exist before returning\n const parsed = JSON.parse(jsonString) as Partial;\n if (!parsed.scene || !parsed.reason) {\n console.error(\"AI JSON missing 'scene' or 'reason':\", jsonString);\n throw new Error(\"AI returned a malformed specification.\");\n }\n\n return new Response(jsonString, {\n headers: { \"Content-Type\": \"application/json\" },\n });\n } catch (error) {\n console.error(\"Generative Art Val Error:\", error);\n return new Response(\n JSON.stringify({\n error: error.message || \"An unknown server error occurred.\",\n }),\n {\n status: 500,\n headers: { \"Content-Type\": \"application/json\" },\n },\n );\n }\n }\n\n // --- GET Request: Serve the HTML frontend ---\n const sourceUrl = `https://${url.hostname}${url.pathname}`;\n return new Response(generateHtml(sourceUrl), {\n headers: { \"Content-Type\": \"text/html\" },\n });\n}","deleted","label",[99],"permalink","https://www.val.town/x/Aperture/Artsy/code/main.ts","extension","actionData","errors"]