Aetherium AI Workflow Engine

\n \n \n

\n \n

Execution Plan

\n \n \n

\n \n

Workflow Execution

\n \n

Workflow Complete

\n \n

\n\n\n\n`;\n}\n\n// ============================================================================\n// CORE ENGINE - WORKFLOW PLANNER\n// ============================================================================\n\nclass WorkflowPlanner {\n private openai: OpenAI;\n\n constructor(openai: OpenAI) {\n this.openai = openai;\n }\n\n async createPlanFromTemplate(\n template: Template,\n context: WorkflowContext,\n ): Promise {\n const filledPrompt = template.promptTemplate.replace(\n /\\{\\{(\\w+)\\}\\}/g,\n (_, key) => context[key.trim()] || `[${key.trim()}]`,\n );\n\n const planningPrompt = `\nYou are an expert workflow architect. Based on the user's goal and a predefined template, refine and detail the execution steps.\n\n**Overall Goal:** ${filledPrompt}\n\n**Template Steps:**\n${JSON.stringify(template.stepDefinitions, null, 2)}\n\n**Your Task:**\nFor each template step, create a specific, actionable, and self-contained prompt for another AI agent to execute. The prompt should incorporate relevant details from the user's context.\n\n**User Context:**\n- Company: ${context.company}\n- Industry: ${context.industry}\n- Overall Goal: ${context.goal}\n${Object.entries(context).map(([k, v]) => `- ${k}: ${v}`).join(\"\\n\")}\n\nRespond STRICTLY with a JSON object containing a single key \"steps\". This key should hold an array of step objects, where each object has \"name\", \"type\", \"dependencies\", \"estimatedTime\", and a \"prompt\" field you have generated.\n${GENERATION_STYLE_GUIDE}\n`;\n\n const response = await this.openai.chat.completions.create({\n model: \"gpt-4o\",\n messages: [\n { role: \"user\", content: planningPrompt },\n ],\n response_format: { type: \"json_object\" },\n });\n\n const planData = JSON.parse(response.choices[0].message.content);\n\n const steps: WorkflowStep[] = planData.steps.map((step: any, idx: number) => ({\n ...step,\n id: `step_${idx + 1}`,\n status: \"pending\",\n metadata: { prompt: step.prompt }, // Nest the generated prompt\n }));\n\n return {\n id: `wf_${Date.now()}`,\n context,\n steps,\n totalEstimatedTime: steps.reduce((sum, s) => sum + s.estimatedTime, 0),\n status: \"draft\",\n createdAt: new Date().toISOString(),\n };\n }\n}\n\n// ============================================================================\n// CORE ENGINE - WORKFLOW EXECUTOR\n// ============================================================================\n\nclass WorkflowExecutor {\n private openai: OpenAI;\n\n constructor(openai: OpenAI) {\n this.openai = openai;\n }\n\n async executeStep(\n step: WorkflowStep,\n planContext: WorkflowContext,\n previousResults: Map,\n ): Promise {\n const dependencyContext = this.buildStepContext(step, previousResults);\n\n const systemMessage = this.getSystemMessageForStepType(step.type, planContext);\n\n const userPrompt = `\n**Execution Context:**\n- Company: ${planContext.company}\n- Industry: ${planContext.industry}\n- Language: ${planContext.language || \"English\"}\n\n**Previous Step Results (for context):**\n${dependencyContext || \"This is the first step.\"}\n\n---\n\n**Your Assigned Task:**\n${step.metadata.prompt}\n\nExecute this task and provide a comprehensive response. If the task requires structured data, respond in JSON format.\n`;\n\n const response = await this.openai.chat.completions.create({\n model: \"gpt-4o\",\n messages: [\n { role: \"system\", content: systemMessage },\n { role: \"user\", content: userPrompt },\n ],\n temperature: 0.5,\n });\n \n const content = response.choices[0].message.content;\n\n return {\n type: step.type,\n content: content,\n };\n }\n\n private buildStepContext(step: WorkflowStep, previousResults: Map): string {\n let context = \"\";\n for (const depId of step.dependencies) {\n const result = previousResults.get(depId);\n if (result) {\n context += `\\n--- Results from ${depId} ---\\n`;\n context += typeof result.content === \"string\" ? result.content : JSON.stringify(result.content, null, 2);\n }\n }\n return context;\n }\n\n private getSystemMessageForStepType(type: WorkflowStep[\"type\"], context: WorkflowContext): string {\n const baseMessages = {\n research:\n \"You are a world-class research analyst. Provide comprehensive, accurate, and well-structured information. Cite sources if possible.\",\n analyze:\n \"You are a senior strategic analyst. Identify key insights, patterns, and actionable recommendations from the provided data. Your analysis must be sharp and conclusive.\",\n generate:\n `You are an expert content creator specializing in the ${context.industry} sector. Produce professional, polished, and compelling content that is ready for publication.`,\n refine:\n \"You are a meticulous editor and quality assurance specialist. Review the provided content, correct any errors, and enhance its clarity, impact, and overall quality.\",\n };\n return `${baseMessages[type]} ${GENERATION_STYLE_GUIDE}`;\n }\n}\n\n// ============================================================================\n// TEMPLATE LIBRARY\n// ============================================================================\n\nconst TEMPLATE_LIBRARY: Template[] = [\n {\n id: \"competitor_analysis\",\n name: \"Competitor Analysis Report\",\n description: \"Deep dive into your competitive landscape with actionable insights.\",\n category: \"strategy\",\n icon: \"🎯\",\n tags: [\"research\", \"strategy\", \"market-analysis\"],\n promptTemplate: \"Analyze competitors in the {{industry}} space for {{company}}\",\n requiredContext: [],\n stepDefinitions: [\n {\n name: \"Identify Key Competitors\",\n type: \"research\",\n dependencies: [],\n estimatedTime: 60,\n metadata: {\n prompt: \"Identify and list the top 3-5 direct and indirect competitors for the company in the specified industry.\",\n }\n },\n {\n name: \"Analyze Competitor Products & Pricing\",\n type: \"analyze\",\n dependencies: [\"step_1\"],\n estimatedTime: 90,\n metadata: {\n prompt: \"For each identified competitor, analyze their core product offerings, key features, and pricing strategies. Present this as a structured comparison.\",\n }\n },\n {\n name: \"Generate SWOT Analysis\",\n type: \"generate\",\n dependencies: [\"step_2\"],\n estimatedTime: 75,\n metadata: {\n prompt: \"Based on the analysis, generate a concise SWOT (Strengths, Weaknesses, Opportunities, Threats) analysis for each competitor relative to our company.\",\n }\n },\n {\n name: \"Synthesize Executive Report\",\n type: \"generate\",\n dependencies: [\"step_3\"],\n estimatedTime: 90,\n metadata: {\n prompt: \"Synthesize all findings into a final executive summary. Highlight key strategic threats and opportunities for our company and provide 3 actionable recommendations.\",\n }\n },\n ],\n estimatedTime: 315,\n successRate: 0.94,\n usageCount: 1247,\n },\n {\n id: 'product_launch',\n name: 'Product Launch Marketing Plan',\n description: 'Go-to-market strategy and initial marketing materials for a new product.',\n category: 'product',\n icon: '🚀',\n tags: ['gtm', 'marketing', 'product'],\n promptTemplate: 'Create a product launch plan for {{productName}} targeting {{targetAudience}}',\n requiredContext: [\n { key: 'productName', label: 'Product Name', placeholder: 'e.g., Fusion Reactor Mini' },\n { key: 'targetAudience', label: 'Target Audience', placeholder: 'e.g., Early-adopter tech homeowners' }\n ],\n stepDefinitions: [\n { name: 'Define Core Messaging', type: 'generate', dependencies: [], estimatedTime: 70, metadata: { prompt: 'Develop a core messaging framework, including a value proposition, a tagline, and three key benefit pillars for the specified product and audience.' } },\n { name: 'Identify Marketing Channels', type: 'analyze', dependencies: ['step_1'], estimatedTime: 60, metadata: { prompt: 'Based on the target audience, identify the top 5 most effective marketing channels for the initial launch phase. Justify each choice.' } },\n { name: 'Create Launch Blog Post', type: 'generate', dependencies: ['step_1', 'step_2'], estimatedTime: 90, metadata: { prompt: 'Write a compelling 600-word blog post announcing the product launch. Incorporate the core messaging and structure it for high engagement.' } },\n { name: 'Draft Social Media Campaign', type: 'generate', dependencies: ['step_1', 'step_3'], estimatedTime: 80, metadata: { prompt: 'Draft a sequence of 5 social media posts (for LinkedIn and Twitter) for the week of the launch. Include text and suggestions for visuals.' } },\n { name: 'Outline Launch Timeline', type: 'generate', dependencies: ['step_2', 'step_4'], estimatedTime: 60, metadata: { prompt: 'Create a high-level Gantt chart timeline for the two weeks leading up to launch day and the week after. Output as a JSON object.' } },\n ],\n estimatedTime: 360,\n successRate: 0.91,\n usageCount: 892,\n },\n];\n\n\n// ============================================================================\n// API & APPLICATION LOGIC\n// ============================================================================\n\nconst app = new Hono();\nconst openai = new OpenAI({ apiKey: Deno.env.get(\"OPENAI_API_KEY\") });\n\napp.get(\"/api/templates\", (c) => {\n return c.json({ templates: TEMPLATE_LIBRARY });\n});\n\napp.post(\"/api/workflow/plan\", async (c) => {\n try {\n const { context, templateId } = await c.req.json();\n if (!context || !templateId) {\n return c.json({ error: \"Context and templateId are required\" }, 400);\n }\n\n const template = TEMPLATE_LIBRARY.find((t) => t.id === templateId);\n if (!template) {\n return c.json({ error: \"Template not found\" }, 404);\n }\n \n const planner = new WorkflowPlanner(openai);\n const plan = await planner.createPlanFromTemplate(template, context);\n\n await blob.setJSON(plan.id, plan);\n\n return c.json({ plan });\n } catch (error) {\n console.error(\"Planning error:\", error);\n return c.json({ error: \"Failed to create plan\", details: error.message }, 500);\n }\n});\n\napp.post(\"/api/workflow/:id/start\", async (c) => {\n const workflowId = c.req.param(\"id\");\n const plan: WorkflowPlan = await blob.getJSON(workflowId);\n\n if (!plan) {\n return c.json({ error: \"Workflow not found\" }, 404);\n }\n\n plan.status = \"approved\";\n await blob.setJSON(workflowId, plan);\n\n return c.json({ success: true, workflowId });\n});\n\napp.get(\"/api/workflow/:id/execute\", (c) => {\n const workflowId = c.req.param(\"id\");\n \n return streamSSE(c, async (stream) => {\n const plan: WorkflowPlan = await blob.getJSON(workflowId);\n if (!plan || plan.status !== 'approved') {\n await stream.writeSSE({ data: JSON.stringify({ type: 'workflow_error', error: 'Workflow not found or not approved.' }) });\n stream.close();\n return;\n }\n\n plan.status = \"running\";\n await blob.setJSON(workflowId, plan);\n \n const executor = new WorkflowExecutor(openai);\n const results = new Map();\n \n try {\n for (const step of plan.steps) {\n await stream.writeSSE({ data: JSON.stringify({ type: 'step_start', step: step.id, name: step.name }) });\n step.status = 'running';\n await blob.setJSON(workflowId, plan);\n \n try {\n const result = await executor.executeStep(step, plan.context, results);\n step.status = 'complete';\n step.result = result;\n results.set(step.id, result);\n await stream.writeSSE({ data: JSON.stringify({ type: 'step_complete', step: step.id, result }) });\n } catch(stepError) {\n console.error(`Error in step ${step.id}:`, stepError);\n step.status = 'error';\n step.error = stepError.message;\n await stream.writeSSE({ data: JSON.stringify({ type: 'step_error', step: step.id, error: stepError.message }) });\n throw new Error(`Execution failed at step: ${step.name}`);\n } finally {\n await blob.setJSON(workflowId, plan);\n }\n }\n \n plan.status = 'complete';\n plan.completedAt = new Date().toISOString();\n await blob.setJSON(workflowId, plan);\n \n await stream.writeSSE({ data: JSON.stringify({ type: 'workflow_complete', workflowId }) });\n } catch(workflowError) {\n plan.status = 'failed';\n await blob.setJSON(workflowId, plan);\n await stream.writeSSE({ data: JSON.stringify({ type: 'workflow_error', error: workflowError.message }) });\n } finally {\n stream.close();\n }\n });\n});\n\nexport default async function(req: Request) {\n const url = new URL(req.url);\n if (url.pathname.startsWith(\"/api/\")) {\n return app.fetch(req);\n }\n const sourceUrl = `https://${url.hostname}${url.pathname}`;\n return new Response(generateHtml(sourceUrl), { headers: { \"Content-Type\": \"text/html\" } });\n}","deleted","label","children",[],[150],"files",[148],"counts",{"_173":32,"_174":175,"_176":32,"_177":178},"pulls","branches",1,"forks","envVars",{"_179":32,"_180":32},"count","emptyCount","contents",[183],{"_102":149,"_95":150,"_151":152,"_153":-5,"_154":136,"_155":135,"_156":157,"_104":184,"_99":159,"_160":161,"_162":163,"_164":22,"_165":-5,"_98":185},["D",1759991968869],[150],"file","upstreamProject","upstreamBranch","emptyEnvVars","isNew","flags",[],"routes/_app.x.$handle.$name._index",{"_93":94,"_95":96,"_97":-7,"_99":-7,"_100":195,"_181":200,"_203":-7,"_204":-5,"_205":-5,"_169":206,"_210":211,"_212":-5,"_213":-5},{"_102":103,"_95":96,"_104":196,"_106":-5,"_107":22,"_108":109,"_110":-5,"_111":-5,"_14":112,"_113":-5,"_114":22,"_115":-5,"_116":32,"_16":117,"_118":119,"_120":119,"_121":122,"_123":122,"_124":122,"_125":119,"_126":119,"_23":197,"_130":198,"_133":199,"_145":22},["D",1759982204756],{"_93":94,"_128":129},{"_93":94,"_132":129},{"_99":135,"_95":43,"_102":136,"_137":103,"_138":-5,"_139":-5,"_140":-5,"_141":-5,"_142":-5,"_143":96,"_144":94},[201],{"_102":149,"_95":150,"_151":152,"_153":-5,"_154":136,"_155":135,"_156":157,"_104":184,"_99":159,"_160":161,"_162":163,"_164":22,"_165":-5,"_98":202},[150],"readme","upstream","pull",[207],{"_102":149,"_95":150,"_151":152,"_153":-5,"_154":136,"_155":135,"_156":157,"_104":158,"_99":159,"_160":161,"_162":163,"_164":22,"_165":-5,"_166":208,"_98":209},[],[150],"previews",[207],"ahead","behind","actionData","errors"]

volt

untitled-5780