Day 15: Building a Multi Platform Social Media Campaign System with goose Recipes
The marketing team needed a coordinated social media push across Instagram, Twitter X, and Facebook. Each platform required a different tone, structure, and content style.
This challenge introduced sub recipes in goose. Instead of writing three separate pieces of content by hand, the goal was to build a reusable automation system. One input. Three platform specific outputs. A single orchestrator recipe coordinating the entire workflow.
The Challenge: Automate Multi Platform Content Generation
The task was to create a four recipe system:
- instagram-post.yaml
- twitter-thread.yaml
- facebook-event.yaml
- social-campaign.yaml (main orchestrator)
Each recipe needed to accept the same core parameters:
- event_name
- event_date
- event_description
- target_audience
- call_to_action
The orchestrator recipe needed to call all three sub recipes and produce a complete campaign package.
The Social Media Campaign System
The completed system generates:
- A fashionable, high impact Instagram caption
- A concise, professional five tweet Twitter X thread
- A warm, family oriented Facebook event description
All content is produced from a single input set and saved to a unified output file. Each recipe was validated, structured with proper YAML frontmatter, and tailored to the communication style of its platform.
Technical Architecture Diagram
Below is a generated text based diagram showing how the recipe system is structured.
┌──────────────────────────────────────────┐
│ Social Campaign System │
└──────────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────┐
│ Main Orchestrator Recipe │
│ social-campaign.yaml │
└──────────────────────────────────────────────┘
│
┌──────────────────────────────────────────────────────────────────────────┐
│ │
▼ ▼
┌──────────────────────────┐ ┌──────────────────────────┐
│ instagram-post.yaml │ │ twitter-thread.yaml │
│ Platform specific caption │ │ Multi tweet thread │
└──────────────────────────┘ └──────────────────────────┘
│
▼
┌──────────────────────────┐
│ facebook-event.yaml │
│ Long form event content │
└──────────────────────────┘
This architecture allows the orchestrator to call each sub recipe independently, aggregate the results, and produce a complete campaign package.
Technical Stack
Below is my stack that powers this system.
| Layer | Technology | Purpose |
|---|---|---|
| Runtime | goose CLI with Recipes extension | Executes recipes and sub recipes |
| Orchestration | social-campaign.yaml | Coordinates multi recipe execution |
| Sub Recipes | instagram, twitter, facebook YAML files | Platform specific content generation |
| Parameters | YAML schema with required fields | Ensures consistent input across all platforms |
| Output | Markdown file | Unified campaign package |
| Reasoning | goose LLM engine | Generates platform appropriate content |
This is a recipe based automation stack designed for repeatable, scalable content generation.
Platform Outputs
The system generated three fully formatted outputs for the Magic Night of Lights and Ice Sculpture Unveiling. Each output reflects the tone and expectations of its platform.
Instagram A high impact caption with strategic hashtags and a polished, visual forward tone.
Twitter X A five tweet thread under 280 characters per tweet, structured for clarity and shareability.
Facebook A long form event description written for families and community oriented audiences.
My Final Thoughts
Day 15 shifted the focus from knowledge engineering to workflow automation. Recipes appear simple at first glance, but orchestrating multiple sub recipes into a cohesive system requires architectural thinking. What stood out in this challenge was the clarity that comes from building reusable automation. Instead of writing content three separate times, I now have a system that will work for every future event with no additional effort.
This is engineering that scales. This is engineering that saves teams time.
This is exactly where goose excels. It rewards structure, clarity, and repeatability.
Day 15 encouraged me to think like a systems designer rather than a content generator, and that shift will matter in the challenges ahead.
This post is part of my Advent of AI journey - AI Engineering: Advent of AI with goose Day 15 of AI engineering challenges.
Follow along for more AI Engineering Adventures with Eri!
Top comments (0)