Step 2: Connect OpenAI Models and Memory

Wire up the AI language models and the session memory used by the agent nodes.

1. Open OpenAI Scoring Model and select the model gpt-4o. Credential Required: Connect your openAiApi credentials.
2. Open OpenAI Content Model and select the model gpt-4.1. Credential Required: Connect your openAiApi credentials.
3. Open Session Memory Buffer and set Session Key to {{ $execution.id }} with Session ID Type set to Custom Key.
4. Confirm Session Memory Buffer is connected to Article Scoring Agent via the ai_memory connection.

Step 3: Set Up Article Scoring and Parsing

Score incoming articles and normalize the AI output into structured fields for routing.

1. Open Article Scoring Agent and set Text to Analyze this article: {{ $json.title }} - {{ $json['content:encoded'] }}.
2. Ensure OpenAI Scoring Model is connected as the language model for Article Scoring Agent.
3. Open Parse Score Payload and keep the provided JavaScript Code to parse the AI JSON response and merge it with the RSS fields.
4. Open Quality Gate Check and set the condition to Number > 4 with Left Value set to {{ $json.quality_score }}.

Step 4: Build Content and Parse Outputs

Generate social content and prepare output fields for publishing and logging.

1. Open Content Builder Agent and set Text to Create social media content for this article: with the full template shown in the node, including {{ $json.title }}, {{ $json.summary }}, {{ $json.content_angle }}, {{ $json.link }}, {{ $json.quality_score }}, and {{ $json.relevance_score }}.
2. Ensure OpenAI Content Model is connected as the language model for Content Builder Agent.
3. Open Parse Content Output and keep the provided JavaScript Code to parse the generated JSON into fields like twitter_thread and linkedin_post.
4. Confirm the execution flow is Quality Gate Check → Content Builder Agent → Parse Content Output.

Step 5: Configure Routing and Output Destinations

Route qualified content to publishing and spreadsheets, and send low-quality items to review or archive.

1. Note the parallel execution: Parse Content Output outputs to both Publish Social Posts and Append Content Sheet in parallel.
2. Open Publish Social Posts and set User to automated-tests, Operation to uploadText, Title to {{ $json.twitter_thread }}, and LinkedIn Title to {{ $json.linkedin_post }}. Set Target LinkedIn Page ID to [YOUR_ID]. Credential Required: Connect your uploadPostApi credentials.
3. Open Append Content Sheet, set Operation to append, and map columns like Article Link to {{ $json.link }} and Twitter Content to {{ $json.twitter_thread[0] }}{{ $json.twitter_thread[1] }}{{ $json.twitter_thread[2] }}{{ $json.twitter_thread[3] }}. Credential Required: Connect your googleSheetsOAuth2Api credentials.
4. Open Secondary Quality Check and set the condition to Number ≥ 5 with Left Value set to {{ $json.quality_score }}.
5. Open Queue Review Sheet with Operation set to append and map fields such as Link to {{ $json.link }} and Content Angle to {{ $json.content_angle }}. Credential Required: Connect your googleSheetsOAuth2Api credentials.
6. Open Archive Records Sheet with Operation set to append and map fields such as Reason to {{ $json.summary }}. Credential Required: Connect your googleSheetsOAuth2Api credentials.

Step 6: Test and Activate Your Workflow

Validate end-to-end processing before enabling the workflow in production.

1. Use RSS Intake Monitor to trigger a manual test run and observe items flowing into Article Scoring Agent and Parse Score Payload.
2. Confirm that Quality Gate Check routes higher-quality items to Content Builder Agent and lower-quality items into Secondary Quality Check for review or archive.
3. Verify that successful runs create rows in Append Content Sheet, and optionally publish via Publish Social Posts with the expected twitter_thread and linkedin_post.
4. Activate the workflow by toggling it to Active after testing confirms outputs are correct.