Build it once, it runs always: n8n and Make templates and clips in Cerebrum

2026-07-28

Producing one clip or post by hand means five open tabs and a downloads folder with files you cannot make sense of a day later. It is slow not because of the generation but because of passing things between steps by hand.

Producing one clip or post by hand means five open tabs and a downloads folder with files you cannot make sense of a day later. It is slow not because of the generation but because of passing things between steps by hand. Below are the two roads that turn "build it once" into "it runs always", with step-by-step instructions for each.

In short, if you are pressed for time

Producing content by hand means five tabs and a downloads folder, and it is slow because of shuffling files, not because of generation.

"Build once, runs always" on the platform is two roads. Schedules, volume and publishing are handled by an n8n or Make workflow. A quality clip for a given task is assembled by Cerebrum or the AI SKILLS connector for Claude.

You do not have to start from a blank page: the catalog holds more than 3,000 ready-made n8n and Make templates. A template is a JSON file: download it, import it, connect your own accounts, switch it on.

Regular templates open on the Creator, Make, N8N and Full plans, advanced ones only on Full, or for a one-off $9, or for 1000 XP. Descriptions are visible to everyone.

A clip from a single task is assembled in Cerebrum, and on the Full plan also in Claude through the connector. The estimate is visible before launch.

There are five types of chain, and a linear one does not always fit.

With clients this is a different price category: a system or pipeline runs $490-1,460 against a median one-off job of about $60.

Five windows and a downloads folder

You opened a tool and wrote the script. Copied it. Opened an image generator, pasted the frame description, waited, downloaded. Opened animation, uploaded the frame, waited, downloaded. Opened narration, generated, downloaded. Opened editing, assembled it all. Then posted it to three social networks.

Five tools, four downloads, four uploads. The generation itself takes minutes, and the evening goes on shuffling files. And tomorrow it is all the same again.

What "build once" means

A chain is a sequence in which one step's result automatically becomes the next step's input. You describe it once and after that it runs by itself. On the platform this idea comes in two forms.

A saved n8n or Make workflow. It takes data (a table row, a new product, an email, a file in a folder), passes it through the steps and publishes or sends the result. An active workflow runs on the n8n or Make server on a schedule or a webhook, with no need for your computer. It stays with you: the same scheme handles one table row or a hundred, and within a month you have your own library of processes.

A clip assembled from one task. You describe the result, and the Gen AI models and the platform's montage pipeline produce the frames, video, voice and edit. There is no schedule here: each clip is a separate task.

Together they cover the whole cycle: n8n or Make takes care of regularity and publishing, Cerebrum assembles the clip. One honest caveat: there is no direct link where n8n calls Cerebrum. An n8n or Make template generates texts, images and video itself through the services connected inside it. Cerebrum is for when a quality edit for your particular task matters.

Road 1. Ready-made n8n and Make templates for content and sales

Every name leads to a catalog card (open the section). "Advanced" means Full, or $9 one-off, or 1000 XP.

A template works through the services named in its card (generators, voiceover, publishing). They are connected with your own keys and paid at their own rates: the platform provides the workflow, not access to third-party services.

How to run an n8n or Make template

Step 1. Open the template card. Read which trigger and which services it needs, and create any missing accounts before importing.

Step 2. Download the JSON file from the card (a workflow for n8n, a blueprint for Make). Links are open on the Creator, Make, N8N and Full plans; advanced templates on Full or through one-off access.

Step 3. Import. In n8n: Workflows, create a workflow, choose Import from File in the menu. In Make: Create a new scenario, the three-dot menu at the bottom of the editor, Import Blueprint.

Step 4. Connect your own credentials in the nodes that need access: Google Sheets, Telegram, OpenAI, social networks. There are no one else's keys in the file.

Step 5. Put in your own data: the table, the folder, the schedule, the prompt texts.

Step 6. Test on one example (Execute workflow in n8n, Run once in Make). Point publishing templates at a test account first.

Step 7. Switch it on: the Active toggle in n8n, enabling the scenario schedule in Make. From then on it runs by itself.

Five types of chain

It helps to know what exists so you do not build a linear chain where another kind is needed. The catalog has an example for each type.

Linear: step by step, idea, text, image, publication. Example: captions from Google Sheets in Make.

Branching: a condition picks the variant, using IF and Switch nodes in n8n and a router in Make. Example: sorting website inquiries, where the inquiry type decides where it goes and what to reply.

Bulk: a set of steps over a list, Split In Batches in n8n and Iterator in Make. Thirty posts from a table in one run; the same idea in the daily short video pipeline.

With a human in the loop: the chain waits for your decision, for example a reply in Telegram or an email, and goes on with what you chose. Example: the cross-posting factory with approval.

With external data: an HTTP request, a table or a database supplies information. Example: LinkedIn auto-posting from a content plan.

Road 2. Ready-made clip pipelines in Cerebrum and the connector

When you need a good clip rather than scheduled publishing, the platform's montage pipeline does the work.

Expert clips. You upload a video where you speak. The pipeline transcribes the speech, asks clarifying questions (call to action, music, what to cut) and assembles a clip with subtitles, headlines, graphics, inserts and sounds. Styles vary, for example a layered cutout collage or an advertising look; you can ask for the plan to be approved before rendering. Modes include a clip without a face, cutting a video of up to 20 minutes into short clips, a website promo, and dubbing into another language. A 9:16 cover and a publishing pack (captions, hashtags) are free.

Clips with your AI character and voice. The character and voice clone you trained in Gen AI are placed into clips: the same face and voice from clip to clip.

Advertising and cinematic clips. Starting frames are generated as photos, a storyboard is built from them, the video model you pick animates each frame, and the scenes are joined with music, voiceover and subtitles. You can attach photos of your own products.

For typical tasks there are free pipeline skills (catalog, the "Content generation" category):

There are more: "Instagram post turnkey", "Product reshoot: all angles", "Article and blog cover", "YouTube thumbnail", "Virtual clothing try-on", "Logo and branding pack", "Instagram analyst".

How to build a clip from one task in Cerebrum

Step 1. Open Cerebrum. Registration is free and gives three days of the Basic plan, no card. Paid generations need money on your balance, and it does not expire.

Step 2. Pick a skill in the chat library or describe the task in plain words: "a 30-second clip from my product photo".

Step 3. Attach a photo or video with the paperclip or by link (file up to 20 MB).

Step 4. The chat shows the estimate: how much will be charged and how much remains on the balance. A paid step does not start without your consent.

Step 5. Confirm. The right models are chosen automatically, and the chat takes the task through the steps, asking clarifying questions where needed.

Step 6. Collect the result in "My generations". Charges follow what was actually done.

Cerebrum has no schedule or auto-launch: each clip is a separate task.

The same through the connector in Claude

The AI SKILLS connector for Claude is available on the Full plan.

Step 1. In Claude open Settings, then Connectors.

Step 2. Click Add custom connector and paste https://aiskills.team/mcp.

Step 3. Sign in with your aiskills.team account.

Step 4. Describe the task in an ordinary chat. Claude picks a pipeline, generates frames, video and voice with Gen AI models and edits the clip. For an expert clip it first asks clarifying questions.

Step 5. Pass your files by link: Google Drive with "Anyone with the link" access, a link to an earlier generation from "My generations", or a direct link. Attachments from the chat do not reach the connector.

What this gives you in client work

The second client costs less than the first. The workflow is set up: for a new client only the accounts and the input data change. You sell the same work for the same money while spending many times less time.

Volume stops being a problem. Thirty posts and three are the same workflow with a different run time.

You have something to show. The workflow diagram is a visible argument: the client sees a process, not a promise. And Cerebrum assembles the clips for them.

Price range. According to our bot's feed, 5,986 orders over four months: a system or pipeline runs $490-1,460 against a median one-off job of about $60. A configured process is the second category, not the first.

On money. In Cerebrum and the connector the price is visible before launch, paid from your balance by what was actually done. In n8n and Make the platform charges nothing for runs, but the services inside a template take their own money: run the workflow on one example first and see what a run costs.

The same principle in ready-made solutions

The same principle applies to every template in the catalog: the card comes with a blueprint file (JSON for n8n, blueprint for Make). You load it, enter your accounts, and five minutes later a system is running that takes a professional weeks to develop, and it can be adapted to a client. How to build scenarios yourself is covered by the course “Automations in Make” (Make plan; three practical tutorials at the end) and the course “Automations in n8n” (N8N plan; it has modules on AI agents and RAG).

A chain you built is your asset. But someone else's finished chain that you can sell after configuring it is an asset that cost you nothing. The catalog holds more than 3,000 templates: before building your own, check whether it already exists. Regular templates open on the Creator, Make, N8N and Full plans, advanced ones only on Full, or for $9, or for 1000 XP. The Make and N8N plans cost $190 a month and include everything in Creator plus the course for their own platform.

Where to start today

One action. Recall what you did last time shuffling files between three tools. Write the steps down in order, a list of four or five lines.

Then find a template in the catalog that covers part of the list and run it as is by the instructions above. Change one block to suit you and save it under your own name: that is already your chain. If the task is about a clip, describe it in Cerebrum.

Building it will take less time than that evening with the downloads folder took, and it will work every time.

What to take on the platform for this task

• Automations in Make — a course on Make scenarios; Make plan (Make and N8N include everything from Creator)
• Automations in n8n — an n8n course including AI agents and RAG; N8N plan
• AI Agents: the era of electrified routine — an express course, $9 or the Creator, Make, N8N, Full plans
• Product campaign turnkey — a conductor skill: key visual, platform adaptations, creatives, a UGC clip; runs through the AI SKILLS connector (Full)
• n8n assistant — a Claude Code skill that designs n8n workflows; advanced (Creator+ / $9 / 1000 XP)
• n8n Self-hosted AI Starter Kit — Open Source: a ready kit for running n8n with your own AI environment on your server; the repository link is from Creator

What to read next

Thirty posts in one run - how to add volume.

Content gets made while you sleep - how to take yourself out of the launch.

Competitor analysis by the numbers - an example of a finished chain.

Make or n8n - when you need a connection builder rather than a chain.

AI agents: what they are and what they pay - a full guide to agents, with market rates.

See also on AI SKILLS

• Make and n8n automation templates
• Cerebrum
• Skills for Claude Code
• Courses and learning