LinkedIn Content Generator from Notion

What This Agent Does

This workflow automatically transforms your raw notes and ideas into polished LinkedIn content on an hourly basis. It monitors a Notion database for content marked as "Ready," then uses AI to generate one main LinkedIn post plus three creative variations—giving you four unique posts from a single set of notes. The entire process runs hands-free, updating your Notion workspace with publication-ready content that maintains your authentic voice while maximizing engagement potential.

Key benefits:

• Save 3-4 hours per week on content creation by automating the ideation-to-draft pipeline
• Multiply your content output by generating four distinct posts from one source
• Maintain consistency with hourly checks ensuring your content queue never runs dry
• Preserve quality through AI-powered writing that adapts to LinkedIn best practices

Perfect for: Content creators, thought leaders, marketing teams, and consultants who need to maintain an active LinkedIn presence without spending hours writing posts daily.

Who Is It For

This workflow is designed for:

• Solo entrepreneurs and consultants who capture ideas throughout the day but struggle to turn them into polished posts
• Marketing teams managing multiple LinkedIn accounts who need scalable content production
• Thought leaders and executives who have valuable insights but limited time for content creation
• Content managers looking to build a content buffer and maintain consistent posting schedules
• Anyone who takes notes in Notion and wants to automatically transform them into LinkedIn-ready content

You'll get the most value if you already use Notion for note-taking and want to establish a systematic content creation process.

Required Integrations

Notion

Why it's needed: Notion serves as both your content input source and output destination. The workflow queries your database for content marked as "Ready," then saves all generated posts back to Notion for review and scheduling.

Setup steps:

1. Create a Notion integration:

   • Visit https://www.notion.so/my-integrations
   • Click "+ New integration"
   • Name it "TaskAGI LinkedIn Content" (or your preference)
   • Select the workspace where your content database lives
   • Click "Submit" to create the integration

2. Copy your Internal Integration Token:

   • After creation, you'll see the "Internal Integration Token"
   • Click "Show" then "Copy" to save this token
   • Keep this secure—it provides access to your Notion workspace

3. Share your database with the integration:

   • Open your Notion content database
   • Click the "•••" menu in the top-right
   • Select "Add connections"
   • Find and select your "TaskAGI LinkedIn Content" integration
   • Click "Confirm"

4. Configure in TaskAGI:

   • Navigate to Settings > Integrations in TaskAGI
   • Click "Add Integration" and select "Notion"
   • Paste your Internal Integration Token
   • Click "Connect" to verify the connection

Database requirements: Your Notion database must include:

• A Status property (Select or Status type) with at least "Ready" and "Running" options
• A Notes or Content property (Text type) containing your raw ideas
• The workflow will create new entries in a separate database for generated posts

Anthropic (Claude)

Why it's needed: Claude powers the main content generation, transforming your notes into a primary LinkedIn post and creating three variations. This workflow uses Claude Sonnet 4.5, which excels at maintaining natural voice and understanding nuanced content requirements.

Setup steps:

1. Create an Anthropic account:

   • Visit https://console.anthropic.com
   • Click "Sign Up" and complete registration
   • Verify your email address

2. Generate an API key:

   • Navigate to "API Keys" in the console
   • Click "Create Key"
   • Name it "TaskAGI LinkedIn Workflow"
   • Copy the API key immediately (it won't be shown again)
   • Store it securely

3. Add credits to your account:

   • Go to "Billing" in the console
   • Add a payment method
   • Purchase credits or set up auto-reload
   • Estimated cost: $0.15-0.30 per workflow run (4 posts generated)

4. Configure in TaskAGI:

   • Go to Settings > Integrations
   • Select "Add Integration" > "Anthropic"
   • Paste your API key
   • Click "Test Connection" to verify
   • Save the integration

OpenAI

Why it's needed: OpenAI's GPT-4o generates the three post idea outlines that serve as prompts for Claude to create variations. This multi-AI approach combines GPT-4o's creative ideation with Claude's natural writing style.

Setup steps:

1. Create an OpenAI account:

   • Visit https://platform.openai.com
   • Click "Sign up" and complete registration
   • Verify your email address

2. Generate an API key:

   • Navigate to "API keys" (https://platform.openai.com/api-keys)
   • Click "Create new secret key"
   • Name it "TaskAGI LinkedIn Content"
   • Copy the key immediately (shown only once)
   • Store securely

3. Add billing information:

   • Go to "Billing" > "Payment methods"
   • Add a payment method
   • Set usage limits if desired (recommended: $10-20/month for this workflow)
   • Estimated cost: $0.05-0.10 per workflow run

4. Configure in TaskAGI:

   • Navigate to Settings > Integrations
   • Click "Add Integration" > "OpenAI"
   • Paste your API key
   • Click "Verify" to test the connection
   • Save the integration

Configuration Steps

Step 1: Configure the Hourly Schedule Trigger

The Hourly Schedule node initiates the workflow automatically.

1. Click on the "Hourly Schedule" node
2. Set the interval to 60 minutes
3. Choose your timezone from the dropdown
4. Optional: Set active hours if you only want the workflow to run during business hours (e.g., 9 AM - 6 PM)
5. Enable the trigger by toggling it "On"

Why hourly? This ensures fresh content is processed regularly without overwhelming your AI usage limits or creating duplicate posts.

Step 2: Set Up Query Notion for Ready Content

The Query Notion for Ready Content node searches your database for content ready to be processed.

1. Select the Notion integration you configured earlier
2. Database ID: Click "Select Database" and choose your content database
3. Filter configuration:
   • Property: Select your "Status" field
   • Condition: equals
   • Value: Ready
4. Sort (optional): Sort by "Created time" ascending to process oldest content first
5. Limit: Set to 1 to process one piece of content per run

Example filter:

{
  "property": "Status",
  "select": {
    "equals": "Ready"
  }
}

Step 3: Configure Check Results Exist

The Check Results Exist node prevents errors when no content is ready.

1. Click the "Check Results Exist" node
2. Set condition type to is not empty
3. Input field: Map to the results from the previous Notion query
4. Path configuration:
   • True path: Continue to content processing
   • False path: Route to "No Content Found" (graceful exit)

This conditional logic ensures the workflow doesn't fail when your content queue is empty.

Step 4: Set Up Extract Content Data

The Extract Content Data function node prepares your content for AI processing.

1. Select "JavaScript" as the function type
2. Input mapping: Map the Notion query results to input
3. Code example:

const page = input[0]; // Get first result
return {
  pageId: page.id,
  notes: page.properties.Notes.rich_text[0]?.plain_text || "",
  title: page.properties.Name.title[0]?.plain_text || "Untitled"
};

4. This extracts the page ID, notes content, and title for downstream nodes

Step 5: Configure Update Status to Running

The Update Status to Running node marks content as being processed to prevent duplicate runs.

1. Select your Notion integration
2. Page ID: Map from the Extract Content Data output: {{4063.pageId}}
3. Properties to update:
   • Property name: Status
   • Value: Running

This immediately updates the status, so even if the workflow fails later, the same content won't be reprocessed.

Step 6: Set Up Claude - Main LinkedIn Post

This is where your primary post gets created.

1. Select your Anthropic integration
2. Model: claude-sonnet-4-5-20250929 (pre-configured)
3. System prompt: (Already configured, but you can customize)

You are a LinkedIn content expert. Based on the notes provided, create an engaging LinkedIn post that:
- Hooks readers in the first line
- Uses short paragraphs and line breaks for readability
- Includes relevant insights or actionable takeaways
- Ends with a question or call-to-action
- Is 150-250 words
- Maintains a professional yet conversational tone

4. User message: Map to your notes: {{4063.notes}}
5. Max tokens: 1000
6. Temperature: 0.7 (balance between creativity and consistency)

Step 7: Configure Save Main Post to Notion

This node stores your generated main post.

1. Select your Notion integration
2. Database ID: Choose your output database (can be the same or different from input)
3. Properties to set:
   • Title/Name: Main Post - {{4063.title}}
   • Content: {{4065.content[0].text}}
   • Type: Main (if you have a Type property)
   • Status: Draft
   • Source Page: {{4063.pageId}} (link back to original notes)

Step 8: Set Up OpenAI - Generate 3 Post Ideas

This node creates three different angles for variation posts.

1. Select your OpenAI integration
2. Model: gpt-4o (pre-configured)
3. System prompt:

Based on the notes provided, create an outline with 3 different angles or perspectives for LinkedIn posts. For each angle, provide:
- A hook/opening line
- Main point or insight
- Call-to-action

Format as JSON array with objects containing: hook, mainPoint, cta

4. User message: {{4063.notes}}
5. Temperature: 0.8 (higher for more creative variations)
6. Max tokens: 800
7. Response format: json_object (ensures structured output)

Step 9: Configure Loop Through Post Ideas

The Loop Through Post Ideas node processes each of the three ideas.

1. Array to loop: Map to OpenAI output: {{4067.choices[0].message.content}}
2. Parse as JSON: Enable this option
3. Loop variable name: postIdea
4. This will execute the next two nodes three times, once for each idea

Step 10: Set Up Claude - Generate Variation Post

This node creates each variation post based on the outlines.

1. Select your Anthropic integration
2. Model: claude-sonnet-4-5-20250929
3. System prompt:

You are a LinkedIn content expert. Based on the outline provided, create an engaging LinkedIn post that:
- Uses the provided hook as inspiration for the opening
- Develops the main point with specific examples or insights
- Incorporates the call-to-action naturally
- Is 150-250 words
- Uses short paragraphs and line breaks
- Maintains a professional yet conversational tone

4. User message:

Original notes: {{4063.notes}}

Outline:
Hook: {{postIdea.hook}}
Main Point: {{postIdea.mainPoint}}
Call-to-action: {{postIdea.cta}}

Create a complete LinkedIn post based on this outline.

5. Max tokens: 1000
6. Temperature: 0.7

Step 11: Configure Save Variation to Notion

This node saves each variation (runs three times in the loop).

1. Select your Notion integration
2. Database ID: Same output database as the main post
3. Properties to set:
   • Title/Name: Variation {{$index + 1}} - {{4063.title}}
   • Content: {{4069.content[0].text}}
   • Type: Variation
   • Status: Draft
   • Source Page: {{4063.pageId}}
   • Variation Number: {{$index + 1}}

Note: $index is a loop variable that counts from 0, so adding 1 gives you variations numbered 1, 2, 3.

Testing Your Agent

Initial Test Run

1. Prepare test content:

   • Open your Notion content database
   • Create a new entry with sample notes (150-300 words of content ideas)
   • Set the Status to "Ready"

2. Run the workflow manually:

   • In TaskAGI, click the "Run Test" button on the workflow
   • Watch the execution flow in real-time
   • Each node will light up as it processes

3. Verify each stage:

   After Query Notion (Node 2):

   • Check execution logs show 1 result found
   • Verify the correct page was retrieved

   After Check Results Exist (Node 3):

   • Confirm the workflow took the true path
   • If it took the false path, check your Status filter

   After Extract Content Data (Node 4):

   • Review the output to ensure notes were extracted correctly
   • Verify pageId, notes, and title are all populated

   After Update Status (Node 5):

   • Check your Notion database
   • The test entry should now show Status: "Running"

   After Main Post Generation (Node 6):

   • Review the generated content in the logs
   • Verify it's 150-250 words and follows LinkedIn best practices
   • Check for proper formatting (line breaks, hook, CTA)

   After Save Main Post (Node 7):

   • Open your output Notion database
   • Confirm the main post was created with correct properties

   After Generate 3 Post Ideas (Node 8):

   • Check the execution logs for the JSON output
   • Verify three distinct ideas were generated
   • Each should have hook, mainPoint, and cta fields

   After Loop Completion (Nodes 9-11):

   • Verify three variation posts were created in Notion
   • Each should be unique and based on different angles
   • Check that all variations link back to the source page

Success Indicators

✅ Complete success looks like:

• Original Notion entry status changed to "Running"
• Four new entries in your output database (1 main + 3 variations)
• All posts are 150-250 words with proper LinkedIn formatting
• Each variation takes a distinctly different angle
• All entries link back to the original source notes
• No error messages in the execution log

Testing Edge Cases

Test with empty notes:

1. Create an entry with Status "Ready" but empty notes field
2. Run the workflow
3. Expected: Should complete but generate generic content or show a graceful error

Test with no ready content:

1. Ensure all entries have Status other than "Ready"
2. Run the workflow
3. Expected: Should exit gracefully at the "No Content Found" node

Test with very long notes:

1. Create an entry with 1000+ words of notes
2. Run the workflow
3. Expected: AI should summarize and create focused posts, not exceed token limits

Troubleshooting

"No results found" but content exists with Status: Ready

Cause: Database connection or filter configuration issue

Solutions:

• Verify the Notion integration has access to the database (check "Connections" in