AI Social Media Caption Agent for Airtable-Powered Content Calendars

Unlock Engaging Social Media Content with this AI Agent

This n8n workflow acts as a sophisticated AI Social Media Caption Agent. It connects directly to your Airtable editorial plan, reads new post briefings, and uses OpenAI's GPT-4o to generate creative and audience-specific social media captions, particularly designed for Instagram. The agent is engineered to understand your content goals and target audience by dynamically fetching background information (like demographics, interests, and preferred tone) from a dedicated Airtable base, ensuring captions are highly relevant and impactful.

Key Features & Benefits

• AI-Powered Caption Generation: Utilizes OpenAI's GPT-4o via a Langchain agent to create compelling and creative social media captions.
• Seamless Airtable Integration: Triggers automatically when new records are added to your Airtable editorial plan, reads briefing information, and posts the generated caption back to the specified Airtable field.
• Dynamic Audience & Tone Adaptation: Employs an Airtable tool to retrieve specific target audience data and communication style guidelines from another Airtable table. This allows the AI to tailor captions for maximum resonance.
• Automated Content Workflow: Streamlines your social media content creation pipeline, taking a post from initial briefing in Airtable to a drafted caption ready for review and publishing.
• Highly Customizable AI Behavior: The AI Agent's instructions (system prompt) are detailed and can be further customized to match your brand voice, specific campaign goals, or content pillars.
• Reduces Manual Effort: Significantly cuts down the time and creative energy spent on manually writing social media posts, freeing you up for more strategic tasks.
• Memory-Equipped: Includes window buffer memory for the Langchain agent, allowing for more contextually aware interactions if the workflow were extended for conversational generation (though primarily used for single-shot caption generation here).

• An n8n instance (Cloud or self-hosted).
• OpenAI API Key with access to a suitable model (e.g., GPT-4o, as configured, or gpt-3.5-turbo).
• Airtable account and API credentials (Personal Access Token is recommended).
• An Airtable base for your editorial plan. Key fields needed: a field for the post briefing (e.g., 'Briefing'), a field to store the AI-generated caption (e.g., 'SoMe_Text_KI'), and a trigger field (e.g., 'created_at' or a status field).
• (Optional but recommended for best results) A separate Airtable table/base for 'Background Info' containing details about your target audience, brand tonality, and preferred wording. The workflow is pre-configured to use a tool to access this.

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure Airtable Credentials: Create or select your Airtable credentials in n8n.
4. 'Airtable Trigger: New Record' Node: a. Assign your Airtable credentials. b. Set the Base ID and Table ID to your main editorial plan in Airtable. c. Choose an appropriate 'Trigger Field' (e.g., created_at for new records, or a specific status field).
5. 'Wait 1 Minute' Node: Adjust timing if necessary. This pause allows for manual entry of the briefing in Airtable after a record is created, before the AI processes it.
6. 'Get Airtable Record Data' Node: a. Assign your Airtable credentials. b. Set the Base ID and Table ID to your main editorial plan (same as the trigger). c. The 'ID' parameter ={{ $json.id }} correctly pulls the ID of the record that initiated the workflow.
7. 'OpenAI Chat Model' Node (under 'AI Agent'): a. Assign your OpenAI API credentials. b. Ensure the 'Model' is set (e.g., gpt-4o).
8. 'Background Info' (Airtable Tool Node under 'AI Agent'): a. Assign your Airtable credentials. b. Set the Base ID and Table ID to your Airtable table containing target audience details and communication style guidelines. c. Verify the 'Record ID' parameter (e.g., reckd97lgylz93Ht5) points to the correct record in this table, or modify it to dynamically fetch relevant info if your setup differs.
9. 'AI Agent' Node: a. Crucially, ensure the text parameter correctly references your briefing field. The template uses ={{ $json['Briefing'] }}. If your Airtable field is named differently (e.g., 'Post Brief') and is nested under 'fields' in the output of 'Get Airtable Record Data', you'll need to change this to something like ={{ $('Get Airtable Record Data').item.json.fields['Post Brief'] }}. b. Review and customize the systemMessage. The current prompt is detailed. Adapt the instructions, especially any references to {{ $json['Briefing/Notizen'] }} or similar, to match your actual briefing field name used in the text parameter. The agent is instructed to use the 'Background Info' tool for audience data.
10. 'Window Buffer Memory' Node (under 'AI Agent'): No specific configuration usually needed for this use case; uses a session key based on the incoming item ID.
11. 'Format Fields' Node: This node prepares the AI's output. It currently maps {{ $json.output }} (from the AI Agent) to a field named SoMe Text. No changes needed unless you rename this internal variable.
12. 'Post Caption into Airtable Record' Node: a. Assign your Airtable credentials. b. Set the Base ID and Table ID to your main editorial plan (same as the trigger). c. Ensure the id column for mapping is ={{ $('Get Airtable Record Data').item.json.id }}. d. Map the AI-generated caption to your desired Airtable field. The template uses SoMe_Text_KI mapped to ={{ $json['SoMe Text'] }} (output from 'Format Fields' node). Adjust SoMe_Text_KI if your Airtable field name is different.
13. Test the workflow with a new record in your Airtable editorial plan.
14. Activate the workflow.