Video Prompt Wizard

\n \n\n

Welcome! Let's create your video prompt.

\n \n

\n \n \n

\n\n

Metadata: The Blueprint of Your Scene

\n \n \n \n

\n \n \n

\n\n

Timeline: The Narrative Arc and Dynamic Elements

\n \n \n \n

\n \n \n

\n\n

Video Prompt Generated!

Your structured video prompt is ready. You can copy the JSON below.

Generated Prompt JSON

\n \n

\n\n

AI is thinking...

\n\n \n\n`;\n}\n\nexport default async function(req: Request) {\n const url = new URL(req.url);\n const action = url.searchParams.get(\"action\");\n\n if (req.method === \"POST\" && action === \"wizard\") {\n try {\n const openai = new OpenAI();\n const body = await req.json();\n const phase = body.phase;\n const currentInput = body.current_input;\n const currentState = body.current_state; // This holds current metadata or currentTimelineSequence\n const existingSequences = body.existing_sequences;\n const currentSequenceNumber = body.current_sequence_number;\n const concept = body.concept; // For full_generate action\n\n let promptMessages;\n let modelToUse = \"gpt-4o\";\n\n if (phase === \"metadata\") {\n promptMessages = [\n { role: \"system\", content: METADATA_PROMPT },\n { role: \"user\", content: `Current user input for metadata: ${currentInput}` },\n ];\n } else if (phase === \"timeline\") {\n promptMessages = [\n { role: \"system\", content: TIMELINE_PROMPT },\n { role: \"user\", content: `Current metadata: ${JSON.stringify(currentState)}` }, // Pass current metadata as a whole\n { role: \"user\", content: `Existing timeline sequences: ${JSON.stringify(existingSequences)}` },\n { role: \"user\", content: `Input for current sequence (${currentSequenceNumber}): ${currentInput}` },\n ];\n } else if (phase === \"full_generate\") {\n if (!concept) {\n throw new Error(\"Concept is required for full prompt generation.\");\n }\n promptMessages = [\n { role: \"system\", content: FULL_GENERATE_PROMPT },\n { role: \"user\", content: `Concept: \"${concept}\"` },\n ];\n } else {\n throw new Error(\"Invalid wizard phase specified.\");\n }\n\n const completion = await openai.chat.completions.create({\n model: modelToUse,\n messages: promptMessages,\n response_format: { type: \"json_object\" },\n temperature: 0.7,\n });\n\n const content = completion.choices[0].message.content;\n const parsedContent: AIResponse = JSON.parse(content);\n\n // Robust validation for critical fields\n if (phase !== \"full_generate\" && (!parsedContent.step || !parsedContent.question)) {\n throw new Error(\"AI returned an incomplete guided response structure.\");\n }\n if (phase === \"full_generate\" && (!parsedContent.generated_metadata || !parsedContent.generated_timeline)) {\n throw new Error(\"AI did not return a complete generated prompt structure.\");\n }\n\n return new Response(JSON.stringify(parsedContent), {\n headers: { \"Content-Type\": \"application/json\" },\n });\n\n } catch (error) {\n console.error(\"AI Wizard Backend Error:\", error);\n return new Response(JSON.stringify({ error: error.message || \"An unknown error occurred during AI processing.\" }), {\n status: 500,\n headers: { \"Content-Type\": \"application/json\" },\n });\n }\n }\n\n // Serve HTML for GET requests\n const sourceUrl = `https://${url.hostname}${url.pathname}`;\n return new Response(generateHtml(sourceUrl), { headers: { \"Content-Type\": \"text/html\" } });\n}","deleted","label",[98],"permalink","https://www.val.town/x/join/director/code/main.tsx","extension","actionData","errors"]