Let your AI assistant post to LinkedIn and Instagram for you

Your own Claude or ChatGPT can draft and schedule LinkedIn and Instagram posts through Kadenza once you've connected your accounts and added Kadenza to your AI app. You work in the AI you already use, and every post waits for your approval. Kadenza gives that AI access to your content and creator research, so you can move from an idea to a post you have read and agreed to publish.

Start with a single post so you can see the whole path. Ask for research, give your AI a real story, and review the draft before you choose when it should go out. You still decide what you're willing to say under your own name. The aim is to make that work easier to carry out while keeping the choices clear at each step.

Connect the accounts you want to post from

In Kadenza, use Configure to connect the LinkedIn and Instagram accounts you want to post from. Then choose a platform for your first draft. If you use more than one account for work, be clear about which one the post belongs to. A personal update and a business post may need different wording, even when they begin with the same idea.

For Instagram, have a photo or video ready for the post you want to make. A caption alone isn't enough to publish there. You can use your own finished image, or work with the image tools in your AI app if you have them. Check the whole post together, because a good caption won't fix an image that says something different from the point you're trying to make.

Add Kadenza to your AI app

Open API & MCP in Kadenza and make a key for the AI app you're connecting. Choose what the key can do, then go to that AI app's connector settings. Add Kadenza using the MCP server URL https://app.kadenza.ai/api/mcp and put your key in the connector's private sign-in settings. Never paste the key into chat. It belongs in the settings used to connect the app.

Use the instructions inside API & MCP for the app you've chosen. If your app can't use the sign-in method shown there, check that before moving on. Don't try to solve it by sharing the key in a conversation. Our MCP guide explains the idea behind the connection, and the docs describe what you can do once you're set up.

Check that your AI can see your content

Start by asking your AI which accounts are connected and what content it can see. Compare its answer with your Kadenza account. That gives you a simple way to check that you're working with your own information. If something is missing, sort out the connection before asking it to schedule a post. A confident answer in chat is only useful when it matches what's actually in your account.

Then ask a question that helps with the post you want to write. For example: "Show me posts from the creators I track that beat their own usual results. Help me find a question I could answer for my audience and show me the posts you used." Read those examples before you pick an idea. Use them to understand what readers care about and where your own experience could help.

Give your AI something only you could say

Once you have a topic, add a story or a view from your work. You could say: "Write a LinkedIn draft using the story below. Keep my facts as they are, use plain words, and ask me if you need a detail I haven't given you." Then share the story. A useful brief gives the AI enough truth to work with and leaves room for questions when something is missing.

Read the draft out loud. Change phrases you wouldn't say, and check that the AI hasn't added a result or a client detail you didn't give it. If a claim needs proof, either add the proof or change the claim. Our guide to writing content that sounds like you goes deeper on this pass. The point is to publish something you'd be happy to explain in a real conversation.

Make the Instagram version fit the picture

If you want to use the same idea on Instagram, show your AI the finished image or video and ask for a caption that fits it. A post about a project lesson might work with an image from the work, while a spoken explanation needs a clear opening you can say naturally. Judge each version on its own. You don't have to use the same wording on both platforms.

Ask to see the full text with the media before you approve it. Check that the image belongs with the story and that any names or details are right. If you change the image, read the caption again. Treat an AI-made image with the same care as any other part of the post. You're approving the finished thing that readers will see, so review it as a whole.

Approve the final version and choose the time

Tell your AI which account should carry the post and give it the day and time you want, including your time zone. Ask it to show those details with the final version before you approve. A clear date matters when you're planning ahead, especially if you're travelling or working with people in another place. Keep the post as a draft until you're happy with both the content and the plan.

Every post waits for your approval. Once you've approved and asked for it to be scheduled, check the entry on your Kadenza calendar. Read the planned time and make sure it's the version you agreed to. A scheduled post is waiting to go out, so don't treat a scheduling confirmation as proof it's already live. If you're unsure what happened, check its status before asking again.

When another tool may fit better

As of September 2026, Buffer offers MCP for Claude and ChatGPT on every plan, including Free, and can put AI-created posts into a queue as drafts or scheduled posts you can review. It may suit you better if you have content ready and mainly want scheduling. As of September 2026, Taplio also connects to Claude and ChatGPT for LinkedIn content and says publishing needs your clear confirmation. It's worth a look if LinkedIn is your focus.

Kadenza costs $117 a month, or $840 a year at the founding price for the first 50 members, while available. Choose it when you also want creator research on LinkedIn, Instagram and YouTube, with guides that can turn interest into email leads. Begin with a post you can review from start to finish. Once you like that way of working, use the same clear brief and approval step for your next one.