Mixed Frameworks

Intermediate ai automation creative

Download mixed.zip

Click to zoom

Mixed Frameworks

Multi-framework demo showcasing OpenAI to ElevenLabs pipelines with local storage for voice synthesis and AI-generated content.

The Mixed Frameworks example demonstrates how Habits can seamlessly orchestrate multiple AI services and frameworks in a single workflow.

What It Does

• OpenAI Integration: Generate text content using GPT models
• ElevenLabs Voice: Convert text to natural-sounding speech
• Local Storage: Save generated assets locally for reuse
• Pipeline Architecture: Chain services together elegantly

This example is perfect for understanding how to combine different AI providers into cohesive workflows using Habits.

Requirements

• OpenAI API key
• ElevenLabs API key

Key Files

stack.yaml

version: "1.0"
workflows:
  - id: text-to-voice-to-s3
    path: ./habit.yaml
    enabled: true
server:
  port: 13000
  host: "0.0.0.0"
  frontend: ./frontend


logging:
  level: info                    # trace, debug, info, warn, error, fatal, none
  outputs: [console]             # console, file, json
  format: text                   # text or json
  colorize: true

habit.yaml

id: text-to-voice-to-s3
name: text-to-voice-to-s3

nodes:
  - id: generate-text
    type: activepieces
    data:
      framework: activepieces
      module: '@activepieces/piece-openai'
      operation: ask_chatgpt
      source: npm
      credentials:
        apiKey: '{{habits.env.OPENAI_API_KEY}}'
      params:
        prompt: 'Write a short 2-sentence motivational quote related to: {{habits.input.prompt}}'
        model: gpt-4o-mini
  - id: text-to-speech
    type: n8n
    data:
      framework: n8n
      module: n8n-nodes-elevenlabs
      operation: text-to-speech
      source: npm
      credentials:
        elevenLabsApi:
          xiApiKey: '{{habits.env.ELEVENLABS_API_KEY}}'
      params:
        resource: speech
        text: '{{generate-text}}'
        voice_id: 21m00Tcm4TlvDq8ikWAM
  - id: save-locally
    type: script
    data:
      label: Save Audio Locally
      framework: script
      source: inline
      params:
        audio: '{{text-to-speech.result.0.base64}}'
        language: deno
        script: |
          export async function main(audio: string | ArrayBuffer) {
            const outputDir = '/tmp/habits-audio';
            await Deno.mkdir(outputDir, { recursive: true });
            
            const filename = `audio-${Date.now()}.mp3`;
            const filepath = `${outputDir}/${filename}`;
            
            // Use Buffer for base64 decoding (works in Node.js environment)
            const buffer = typeof audio === 'string' 
              ? Buffer.from(audio, 'base64')
              : new Uint8Array(audio);
            
            await Deno.writeFile(filepath, buffer);
            
            const stats = await Deno.stat(filepath);
            return { 
              success: true, 
              filepath,
              filename,
              size: stats.size,
              modifiedAt: stats.mtime?.toISOString()
            };
          }
edges:
  - source: generate-text
    target: text-to-speech
  - source: text-to-speech
    target: save-locally

.env.example

HABITS_OPENAPI_ENABLED=true
HABITS_MANAGE_ENABLED=true

OPENAI_API_KEY=sk-proj-key12-here34-makesuretocopyitall56
ELEVENLABS_API_KEY=

Quick Start

Run directly using Cortex package, recommended for production runs, does not inlcude base or extra depdencies.

# First, download the example files
npx @ha-bits/cortex@latest server --config ./mixed/stack.yaml

Download mixed.zip