Social Media Manager AI for n8n manual

What it does

• Starts the workflow with a Schedule Trigger set to the cron expression 0 8 * * *.
• Reads range A:D from the Google Sheet specified in the Leggi Topics da Google Sheets node.
• Sends every item read to Genera Post con GPT-4, an OpenAI chat node set to the gpt-4o model.
• Uses incoming values named platform, topic, and tone in the prompt.
• Sends the generated text in parallel to Pubblica su Twitter/X and Pubblica su LinkedIn.
• Connects the output path from Pubblica su Twitter/X to Salva Risultati and then Notifica Telegram.
• Keeps the YOUR_SHEET_ID and YOUR_CHAT_ID placeholders in the JSON, which you must replace before use.
• Contains seven nodes and no pinned data or credentials in the verified file.

What it does not do

• It does not enforce a three-post or three-row daily limit in the JSON.
• It does not include a human approval step before automatic publishing.
• It does not include publishing nodes for Facebook, Instagram, or social networks other than X and LinkedIn.
• It does not connect the Pubblica su LinkedIn branch to result saving or the Telegram notification.
• It does not include a Google Sheet template or create accounts, OAuth apps, Telegram bots, or API keys for you.
• It does not verify that generated text meets platform policies, rights requirements, consent requirements, or legal obligations.

Requirements

• n8n version 1.0 or later: Use a cloud or self-hosted n8n instance that can import the JSON file and includes the nodes used by the workflow.
• Package files: You need n8n-social-media-manager.json and, for reference, the extracted README.md and README.en.md files.
• Google Sheets and n8n authorization: You need a sheet that n8n can read and edit, plus a Google Sheets credential authorized both to read and append rows.
• OpenAI and an n8n credential: The delivered file uses an OpenAI chat node with the gpt-4o model. You need a valid credential for your n8n setup; third-party services may be paid, so check the provider’s price list.
• Authorized X and LinkedIn accounts: You need credentials that the respective n8n nodes can use to publish. Check requirements, permissions, and availability in the provider and n8n documentation.
• Telegram bot and chat ID: For Notifica Telegram, you need a Telegram bot configured in n8n and the recipient chat ID. You can create the bot using Telegram’s standard BotFather procedure.
• Browser and administrative access to your accounts: These are needed to connect credentials and authorize integrations in your n8n instance.

Step-by-step setup

1. Extract the ZIP to a working folder. Download the package and keep the three extracted files together. Do not rename the JSON before importing it so it is easy to identify.
2. Read the delivered guides. Read README.md for the Italian guide or README.en.md for the English guide. The JSON workflow has Italian node labels and an Italian generation prompt, although you can edit that text in your own n8n.
3. Prepare the Google Sheet. Create or use a sheet that n8n can access. The reading node uses range A:D. The AI prompt requires incoming values named platform, topic, and tone; with standard header-based sheet handling, use these headers for the columns that should feed the prompt. The fourth field in range A:D is not referenced by the delivered prompt.
4. Collect the identifiers to enter. Copy your Google Sheet ID and identify the destination Telegram chat ID. Do not place tokens, keys, or credential values in public documents.
5. Create or connect credentials in n8n. In your n8n instance, connect the credentials required by the Google Sheets, OpenAI, Twitter, LinkedIn, and Telegram nodes. Follow the normal authorization screens in your n8n and the providers; the package contains no reusable credentials.
6. Import the workflow. In n8n, open Workflows, choose Import from File, and select n8n-social-media-manager.json. The imported workflow is named DigitalHub — Social Media Manager AI.
7. Assign credentials to the nodes. Open nodes showing warnings and connect the appropriate credentials. In particular, check Leggi Topics da Google Sheets, Genera Post con GPT-4, Pubblica su Twitter/X, Pubblica su LinkedIn, Salva Risultati, and Notifica Telegram.
8. Replace the file placeholders. In Leggi Topics da Google Sheets and Salva Risultati, replace YOUR_SHEET_ID with your actual sheet ID. In Notifica Telegram, replace YOUR_CHAT_ID with the selected chat ID. Keep ranges A:D and F:H only if they match your sheet layout.
9. Review the schedule and prompt. The trigger uses 0 8 * * *, which runs daily at 08:00 according to the time zone configured in your n8n instance. In Genera Post con GPT-4, you can review the two delivered messages to align voice, topics, and hashtags with your brand. This edit does not add human review before publishing.
10. Run a controlled test, then activate. First check the sheet rows and social destinations. Run Execute Workflow in test mode, inspect inputs and outputs node by node, and verify whether the Salva Risultati and Notifica Telegram expressions receive the expected data. Only when the test is correct, turn on Active. Activation lets the scheduled trigger publish without a built-in manual approval step.

Daily use

Keep source data orderly

Before the scheduled time, check that rows read from A:D contain correct platform, topic, and tone values. The workflow does not filter or limit the items in its read range to three.

Treat each execution as direct publishing

Once activated, the workflow has no draft or approval node. Use only topics you are authorized to publish and recheck the schedule before leaving the flow active.

Review executions in n8n

After a test or scheduled start, open the execution and inspect each node’s input and output. This is especially important for the branch after Pubblica su Twitter/X, which feeds Salva Risultati and Notifica Telegram.

Review AI content before expanding use

Assess accuracy, tone, hashtags, brand references, and channel fit. Generated text can need correction and does not replace editorial review.

Update credentials securely

When a connection expires or changes, update the credential in n8n instead of storing tokens in the Google Sheet, prompt, or publicly shared workflow copies.

Practical cases

Small online shop with a topic calendar

A small online shop keeps topics, desired tone, and platform notes for the day in its own sheet.

1. Create columns that produce the platform, topic, and tone fields.
2. Enter one test row about a topic the shop is authorized to communicate.
3. Connect authorized X and LinkedIn accounts and run Execute Workflow.
4. Inspect the generated text and the node outcomes before activating the schedule.
5. Keep the sheet limited to content that has already been internally reviewed.

The shop has a configured flow that starts from its own sheet and makes generation and publishing attempts inspectable in n8n executions.

Consultant with recurring themes

A consultant organizes educational topics and desired tone in a Google Sheet for communication on authorized profiles.

1. Enter informational topics the consultant has checked.
2. State the desired tone in each sheet item.
3. Keep only credentials for accounts the consultant directly manages active.
4. Run a test and check whether the read items are the intended ones.
5. Activate the trigger only after confirming content, destinations, and timing.

The consultant has a repeatable process based on a topic list without delegating content accuracy or suitability checks to the workflow.

Marketing team performing a technical check

A team uses the workflow to test the chain between a sheet, generation, publishing, and notification in an environment with controlled accounts.

1. Prepare non-sensitive test rows in range A:D.
2. Set sheetId and chatId to authorized team resources.
3. Run the workflow manually and open every node in the execution.
4. Pay particular attention to the inputs available to Salva Risultati and Notifica Telegram after the X node.
5. Document any local changes before enabling scheduled execution.

The team obtains technical evidence of how the imported file behaves and can decide whether its own setup needs adjustments before scheduled use.

Tips

• Start with a single test row in the sheet to limit publishing during the first verification.
• Use platform values consistently with the destinations in this workflow, because the delivered file still sends output to both the X and LinkedIn nodes.
• Keep topics specific and verifiable: the generator only receives values present in the incoming item and in the delivered prompt.
• Use tone as a concrete editorial instruction, such as professional, informative, or conversational, and inspect the result you get.
• Check the n8n instance time zone before treating 08:00 as your intended local time.
• Do not assume the count is three: the JSON reads A:D without a node that selects exactly three rows.
• Keep the workflow inactive until you have completed at least one test with authorized data and accounts.
• Explicitly check the LinkedIn branch in test results, because the delivered file does not continue it to Salva Risultati or Notifica Telegram.
• If you change ranges, prompts, or connections, run a complete test again before reactivating the flow.

Troubleshooting

The workflow does not start at the expected time.

The workflow may be inactive, the cron expression may have been changed, or the instance time zone may not be the expected one. Check that Active is enabled, review cronExpression, and check your n8n instance time zone. Run a manual test first to separate a scheduling issue from a node issue.

The Google Sheets node does not read data.

The credential may lack access to the sheet, sheetId may still contain the placeholder, or range A:D may not match the sheet in use. Reconnect an authorized credential, replace YOUR_SHEET_ID, and confirm in Google that the sheet and range are accessible.

platform, topic, or tone is blank in the AI node.

The item returned by the reading node does not contain the fields referenced by the prompt expressions. Open the Leggi Topics da Google Sheets output in a test execution and make the read produce platform, topic, and tone fields. With standard header handling, check the column names.

Genera Post con GPT-4 returns a credential or model error.

The OpenAI credential may be missing, invalid, or not have availability for gpt-4o in the setup in use. Open the node, select a valid credential, and verify model availability with the provider and in your n8n version. Do not place API keys in the prompt or sheet.

X or LinkedIn does not publish.

The credential, account permissions, provider policies, or data sent by the node may not be accepted. Check the credential and execution error in the specific node. Confirm with the provider that the account and authorized app can publish, then review the text before testing again.

No row appears in Salva Risultati or the Telegram notification fails.

In the delivered file, these nodes receive the branch after Pubblica su Twitter/X, while their expressions refer to platform and choices[0].message.content. The data available after the X node may need checking in your test. Open the Pubblica su Twitter/X, Salva Risultati, and Notifica Telegram outputs in the same execution. Check the actual available fields before changing any mappings, then test again with one row.

LinkedIn publishes but no Telegram notification arrives.

In the JSON, Pubblica su LinkedIn has no outgoing connection to Salva Risultati or Notifica Telegram. Treat the delivered notification as connected only to the path after Pubblica su Twitter/X. Review the LinkedIn branch in n8n execution history.

The workflow processes more items than wanted.

The node reads A:D and the file has no node that limits items to three. Reduce or control the rows available in the range before execution and test with one row. If you change the workflow to add a limit, test that as a local customization.

Limits to know

• The delivered guide mentions three posts per day, but the verified JSON has no three-row limit: processing depends on the items read from range A:D.
• Publishing is configured directly in the X and LinkedIn nodes, and the file contains no approval, moderation, or editorial queue.
• The LinkedIn branch ends at the publishing node; only the output path from X connects to Salva Risultati and Notifica Telegram.
• The workflow includes X and LinkedIn references, not Facebook, Instagram, or other channels mentioned only as examples in the README guide.
• AI-generated content must always be reviewed by a person before publication, including accuracy, tone, rights, hashtags, and platform policies.
• Respect service terms, account permissions, privacy rules, and data-protection requirements. Do not put unnecessary personal data in the sheet or prompt, and protect credentials.
• Third-party accounts, APIs, and features may require paid plans or change over time; check each provider’s price list, requirements, and official documentation.

Frequently asked questions

Does the workflow publish exactly three posts every day?

No, not based on the verified JSON. The README describes three posts per day, but the Google Sheets node reads A:D and there is no node that selects or limits items to three. The number depends on the data read.

Which social networks does the delivered file publish to?

The JSON contains a Pubblica su Twitter/X node and a Pubblica su LinkedIn node. Both receive the Genera Post con GPT-4 output in parallel.

Do I need to create a new Google Sheet?

No template is included. You can use an existing sheet or create one, as long as n8n can read it and append rows and its data produces the platform, topic, and tone fields used in the prompt.

What must I change before the first test?

Connect credentials, replace YOUR_SHEET_ID in both Google Sheets nodes and YOUR_CHAT_ID in the Telegram node, then check the sheet range, prompt, and schedule.

Does the Telegram notification also confirm LinkedIn publishing?

Not necessarily. In the delivered file, the LinkedIn node is not connected to Salva Risultati or Notifica Telegram; these nodes follow only the path after Pubblica su Twitter/X.

Can I change voice, topics, and hashtags?

Yes. The README says to edit the AI node prompt. Genera Post con GPT-4 contains the two messages to review and uses the platform, topic, and tone values. Test every change before activation.

Does the package contain an API key or credentials I can use?

No. The verified file has no node credentials and uses placeholder identifiers. You must use your own credentials and store them in n8n’s credential manager.