14-node n8n workflow using GPT-4, GoToHuman, and Blotato for AI-generated LinkedIn posts with human approval gates.
A 14-node content automation system that solves the fundamental problem with AI-generated social media: quality control. This workflow reads from a content calendar, generates LinkedIn posts with GPT-4's "Think" creativity tool, and then stops. It does not post automatically. It sends the draft through GoToHuman for professional review and only publishes to LinkedIn via Blotato after a human approves it. This is human-in-the-loop automation done right -- speed where it helps, judgment where it matters.
| Nodes | Category | Complexity | Key Integrations |
|---|---|---|---|
| 14 | Marketing & Social | Intermediate | OpenAI (GPT-4), GoToHuman, Blotato (LinkedIn), Google Sheets |
30-45 minutes including GoToHuman reviewer setup, Blotato LinkedIn connection, content calendar template creation, and credential configuration.
Every day, this workflow checks a Google Sheets content calendar for today's topic. If a topic is scheduled, GPT-4 generates a LinkedIn post using the "Think" tool -- a specialized reasoning step that forces the AI to consider audience, tone, hooks, and engagement patterns before writing. The result is a draft that reads like it was crafted by a professional copywriter, not assembled by a template engine.
The draft is saved to Google Sheets and simultaneously sent to GoToHuman, which routes it to a designated reviewer via their preferred channel. The reviewer can approve, reject, or request edits. Only after explicit approval does the workflow fire an HTTP request to Blotato, which publishes the post to LinkedIn.
This design means you get daily content generation without the risk of unreviewed AI output going live. The content calendar drives consistency, the AI drives speed, and the human drives quality. All three work together without stepping on each other.
Triggers
AI & Content Generation
Approval & Publishing
Data & Logic
Daily Schedule Trigger -- The workflow fires every morning at a configured time. A manual trigger is also available for on-demand generation outside the regular schedule.
Content Calendar Lookup -- Google Sheets is queried for today's date. The matching row contains the topic, target audience notes, key messages, and any specific instructions for tone or format.
AI Content Generation -- The AI Agent receives the topic and context. Before writing, the Think tool forces a structured reasoning step: it considers the audience, identifies the strongest hook, plans the post structure, and evaluates engagement patterns. Only after this deliberation does GPT-4 produce the actual post text.
Draft Storage -- The generated post is saved to Google Sheets as a draft with metadata including the generation timestamp, the topic it was based on, and the AI's reasoning notes from the Think step.
Human Review via GoToHuman -- The draft is sent to GoToHuman, which routes it to a designated reviewer. The reviewer sees the full post, the original topic brief, and the AI's reasoning. They can approve, reject with notes, or request specific edits.
Approval Filter -- A Filter node checks the GoToHuman response. Rejected posts are logged and the workflow exits. Approved posts continue to the publishing step.
LinkedIn Publishing -- An HTTP Request node sends the approved post to Blotato's API, which handles the LinkedIn posting. The published URL is logged back to Google Sheets, completing the audit trail.
GoToHuman integration for professional approval workflows. Unlike Slack-based "approve/reject" hacks, GoToHuman is purpose-built for human-in-the-loop AI workflows. It provides clean review interfaces, edit capabilities, and audit trails that make content approval feel professional rather than improvised.
The "Think" tool for AI creativity. The Think tool forces GPT-4 to reason through the creative process before generating output. It considers hooks, audience psychology, and post structure deliberately rather than generating text stream-of-consciousness. This produces noticeably higher quality first drafts.
Content calendar driven automation. The workflow does not generate random content -- it follows a planned editorial calendar. This means the AI operates within strategic constraints set by the marketing team, maintaining brand consistency and campaign alignment.
Tip: Key Techniques
Human-in-the-Loop Pattern -- The most effective AI content workflows include a human approval gate before publication. GoToHuman provides a clean implementation of this pattern that scales across teams and maintains audit trails.
Think Tool for Structured Creativity -- The Think tool adds a deliberation step before generation. Rather than hoping GPT-4 produces good output on the first attempt, you force it to plan the output structure, which consistently improves quality.
Content Calendar as Input -- Driving AI content generation from a pre-planned calendar ensures strategic alignment. The AI handles execution, the human handles strategy. This separation of concerns is what makes sustainable content automation possible.
The workflow.json file in this folder can be imported directly into your n8n instance via Settings > Import Workflow.
I build production n8n and Cloudflare automation for teams — the same engineering behind HarperFlow. Fixed-price, escrow-protected, US-based.