What does my agent get that ChatGPT or Claude alone doesn't have?

Your history. A model on its own doesn't know how you write, what you've said before, which of your posts worked or who engaged with them, so it writes like every other agent and fills gaps with invented details. Connected to ContentIn, it writes from your voice profile and your saved stories and numbers, reads your per-post results, sees the people engaging with you scored against your ideal customer (Pro), and posts through LinkedIn's official API after you confirm. The result is posts that sound like you, more of what performs, and leads you can act on.

How do I use it in n8n?

Create an API key in ContentIn under Profiles, Connections, API keys (Growth or Pro). In n8n, add an MCP Client node, set the transport to HTTP Streamable, use the server URL as the endpoint and add a Header Auth credential with the name X-MCP-Key and your key as the value. The tools then show up for your workflow or AI Agent node. Each key acts for one ContentIn profile.

Can an automation publish without me?

No. Scheduling and publishing are two-step: the tool returns the exact text and a confirmation token, and the post only goes out once a person has confirmed that text. Automations are good for gathering analytics and leads, saving material and producing drafts; the last step stays with you.

Is it LinkedIn-safe? Does it use the official API?

Yes. Posting and post analytics go through LinkedIn's official API for the LinkedIn account you connected to ContentIn. Your agent never logs into your LinkedIn account, and your LinkedIn password and cookies never reach ContentIn or your agent, so there is none of the session automation that gets accounts restricted.

Do I need to write prompts or configure the tools?

No. Once connected, you ask in plain language and your agent picks the right tool. There's no system prompt to write. If you build your own agent or workflow, the tool descriptions come with the server.

I run several profiles or an agency. How does that work?

Each connection or API key is bound to one LinkedIn profile, which you pick when you connect. Agencies connect one per client profile, so each client's voice, material and queue stay separate.

Are the Coming next items available?

No. Post library search, draft check, webhooks and the Agent + Leads plan are not available yet. We build them in the order people ask for them; the “I want this” buttons on this page are how we count that.

What happens to my data?

A connected agent can only reach the one ContentIn profile you approved: its posts, drafts, ideas, analytics, comments and leads. It can't see your other profiles, your billing details or your LinkedIn password. You can see and revoke every connection and API key in ContentIn under Profiles, Connections, and revoking takes effect immediately. What your agent does with data it reads is governed by that agent's own terms.

For people who run AI agents and automations

Make your AI agent sound like you on LinkedIn

Your agent can already write. On its own it writes like everyone else's agent: no memory of you, no idea what worked, no idea who is reading. Connect it to ContentIn and it writes from your voice and your own stories, repeats what performed on your account, shows you which buyers engaged, and posts through LinkedIn's official API only after you confirm the text.

Start your 7-day trial Read the setup guide

Runs on the Growth plan. Card required for the trial. Sign-up asks for your LinkedIn profile first.

Works with the agents and automations you already run

Claude
ChatGPT
Claude Code
Gemini
OpenClaw
n8n

Claude, ChatGPT and Claude Code connect with the server URL and a ContentIn sign-in. Gemini connects through Gemini CLI. OpenClaw and n8n connect with the same URL and an API key, so a workflow can run without anyone opening a browser. Also works in Cursor and VS Code.

Live today

What changes when your agent is connected

You keep the agent you like. ContentIn gives it the four things a model can't know on its own, so the posts it helps you ship sound like you, perform better and bring in people you can talk to.

Your voice and your own material

ContentIn learns your voice from the posts you've already published and keeps a bank of your stories, opinions and numbers. Before it writes, your agent reads your voice, your rules, your saved material and your best posts, so whatever model you use (Claude, ChatGPT, Gemini) can write the post itself. Or it asks ContentIn to write the draft. It saves new material to the bank from any conversation or workflow.

The outcome: You keep your favourite AI and it writes like you, from your real stories and numbers instead of invented ones, and stops sounding like every other agent. Less rewriting before you post.

get_voice_contextwrite_post_in_my_voicecapture_substance

What actually worked on your account

Your drafts, scheduled and published posts, with impressions, members reached, reactions, comments and shares per post, from LinkedIn's official API.

The outcome: Your agent picks topics and formats from your own results instead of guessing, so you post more of what performs for you.

list_postsget_post_analytics

The people who engaged, scored against your ideal customer

Everyone who reacted to, commented on or shared your posts, with ContentIn's fit score against your ideal customer, lead status and the posts that pulled them in. Plus the comments you left on other people's posts and what they got.

The outcome: Your posts turn into a short list of buyers to talk to, and you can see which posts attract them.

list_leadslist_my_comments

Pro plan.

A safe last mile

Posts go out through LinkedIn's official API, with up to five images. Before anything is scheduled or published, your agent has to show you the exact text and get your yes. Your agent never logs into your LinkedIn account and never touches your password or cookies.

The outcome: No account restrictions from session or browser automation, and nothing goes out that you haven't read.

schedule_postpublish_postattach_images

ContentIn also has its own idea and rewrite tools (generate_ideas, repurpose_post). Your agent can use them when it helps; it stays in charge.

The same agent, with and without ContentIn

Your agent on its own

Writes in the same polished tone as every other agent.
Fills gaps with plausible stories and numbers you never gave it.
Doesn't know which of your posts worked, so it guesses at topics.
Can't tell you who read your posts or whether they could buy from you.
Posting means copy-paste, or browser automation that risks your account.

Your agent connected to ContentIn

Reads your voice, your rules, your saved material and your best posts before it writes.
Uses the stories, opinions and numbers you saved, and saves new ones as you talk.
Reads the results of your own posts before suggesting the next one.
Lists the people who engaged, scored against your ideal customer (Pro).
Schedules or publishes through LinkedIn's official API after you confirm the text.

Use it in automations

Put your LinkedIn into n8n or OpenClaw workflows

In n8n, add an MCP Client node set to HTTP Streamable, point it at the server URL and send your API key in the X-MCP-Key header. OpenClaw takes the same URL and header. Three workflows that run on the tools that are live today:

Monday LinkedIn report

1.A weekly schedule trigger starts the workflow.
2.It pulls last week's posts and their analytics, and the new people who engaged.
3.An AI step summarises what worked and who to follow up with, and posts it to Slack or your CRM.

list_postsget_post_analyticslist_leads

Why it pays off: You start the week knowing what to post more of and who to talk to, without opening LinkedIn.

The lead list needs Pro.

Note or voice memo to draft

1.A note, or a transcribed voice memo, lands from another app.
2.The workflow saves it to your material bank, then asks ContentIn for a draft in your voice.
3.The draft waits in ContentIn. You read it and confirm before it's scheduled.

capture_substancewrite_post_in_my_voice

Why it pays off: Ideas you'd have forgotten become drafts in your voice, built from what you actually said.

New leads into your CRM

1.Once a day, the workflow asks for the people who engaged with your latest posts.
2.It keeps the ones with a high fit score for your ideal customer.
3.It adds them to your CRM or a sheet, with the post that brought them in.

list_leads

Why it pays off: Engagement turns into a follow-up list you can work, not a notification you scroll past.

Needs Pro. Today the workflow checks on a schedule; pushing new leads the moment they appear needs Webhooks, which is coming next and not available yet.

API keys are created in ContentIn under Profiles, Connections, API keys (Growth and Pro). Each key acts for one profile. The steps are in the setup guide.

Twelve tools live today, four coming next

Your agent picks the tool from what you ask. You never call them by name. Each connection is bound to one LinkedIn profile; agencies connect one per client profile.

list_postsYour drafts, scheduled and published posts.From Growth
get_post_analyticsImpressions, reach, reactions, comments and shares per post.From Growth
list_leadsPeople who engaged with your posts, with ICP fit score and lead status.Pro
list_my_commentsThe comments you left on other people's posts, and what they got.Pro
get_voice_contextYour voice, rules, saved material and best posts, served to your agent before it writes.From Growth
write_post_in_my_voiceA draft written in your voice, from your saved material.From Growth
generate_ideasPost ideas from your topics, audience and what has resonated.From Growth
repurpose_postRework a post that landed from a fresh angle.From Growth
capture_substanceSave a story, opinion or number to your material bank.From Growth
attach_imagesAttach up to five images to a draft.From Growth
schedule_postSchedule a post. Shows the exact text and waits for your yes.From Growth
publish_postPublish now. Shows the exact text and waits for your yes.From Growth
Search what's working on LinkedIn nowAbout 60,000 high-performing LinkedIn posts, updated daily, searchable by topic.Coming next
Draft checkVoice match, AI tells, repeats of old posts and unbacked claims, checked before posting.Coming next
WebhooksNew ICP-fit lead, lead engaged again, post published, analytics ready, pushed to your tools.Coming next
Agent + Leads planICP-scored engagers into your CRM, plus comment drafts you approve.Coming next

Rows marked “Coming next” are not available yet. Details below.

Setup today

Three steps from sign-up to your agent working with your LinkedIn

Drop your LinkedIn profile
Sign up and paste your LinkedIn profile. ContentIn reads your posts and learns your voice from them.
Connect your agent with one URL
In Claude, ChatGPT or Claude Code, add this URL, sign in with ContentIn and pick the profile. No key to copy. For Gemini CLI, OpenClaw or n8n, create an API key in ContentIn and send it in the X-MCP-Key header.
https://mcp.contentin.io/mcp-server
Ask it something real
“Which of my last ten posts did best, who engaged with it, and what should I write next?” Your agent picks the tools; you confirm anything before it goes to LinkedIn.

Per-client steps for Claude, ChatGPT and Claude Code (plus Cursor and VS Code), and API keys for Gemini CLI, OpenClaw, n8n and other agents that can't open a browser, are in the setup guide.

Not available yet

Coming next

None of this is live. We build in the order people ask for it, so if one of these would change what your agent can do for you, press “I want this”.

Not available yet
Search what's working on LinkedIn now
What it is
A library of about 60,000 high-performing LinkedIn posts, updated daily, searchable by topic, with each post's hook and format.
What your agent can do
Before it writes, your agent looks up what's performing on your topic right now and bases the angle, hook and format on it.
Why it's a better outcome
Your posts start from angles that are proven on LinkedIn this month, not from the model's guess.
Not available yet
Draft check
What it is
Send any draft and get back a voice match score, AI tells, repeats of your old posts, and claims your material doesn't back.
What your agent can do
Run every draft through a quality gate before it's scheduled, whoever or whatever wrote it.
Why it's a better outcome
Fewer posts that sound generated, repeat yourself or say something you can't stand behind.
Not available yet
Webhooks
What it is
Events pushed to n8n, Zapier or your CRM: a new lead that fits your ideal customer, a lead engaged again, a post was published, analytics are ready.
What your agent can do
Your automations react the moment something happens instead of checking on a schedule.
Why it's a better outcome
You follow up with a buyer while your post is still fresh in their mind.
Not available yet
Agent + Leads plan
What it is
A plan built for agent users: people who engaged, scored against your ideal customer, sent to your CRM, plus comment drafts you approve before anything is posted.
What your agent can do
Keep your pipeline filled from LinkedIn engagement and draft replies, with you approving every comment.
Why it's a better outcome
More conversations with the right people, without spending your day in the LinkedIn feed.

How early access works

Early access runs on the ContentIn Growth plan (MCP included); leads tools are on Pro. There is no separate agent plan today.

Growth

$39 a month

or $31 a month billed yearly

The MCP connection and API keys with ten of the twelve tools: your posts and analytics, your voice and material for your agent, writing in your voice, your material bank, images, scheduling and publishing.

Start with Growth

Pro

$59 a month

or $48 a month billed yearly

Everything in Growth, plus the leads tools: who engaged with your posts, scored against your ideal customer, and the comments you've left.

Start with Pro

7-day trial. A card is required to start it; cancel before day 7 and you're not charged.
Sign-up asks for your LinkedIn profile first, so ContentIn can learn your voice from your posts. You choose Growth or Pro when you start the trial.
Then connect your agent with https://mcp.contentin.io/mcp-server.

Frequently asked questions

What does my agent get that ChatGPT or Claude alone doesn't have?: Your history. A model on its own doesn't know how you write, what you've said before, which of your posts worked or who engaged with them, so it writes like every other agent and fills gaps with invented details. Connected to ContentIn, it writes from your voice profile and your saved stories and numbers, reads your per-post results, sees the people engaging with you scored against your ideal customer (Pro), and posts through LinkedIn's official API after you confirm. The result is posts that sound like you, more of what performs, and leads you can act on.
Which AI agents and automation tools does it work with?: Claude, ChatGPT and Claude Code connect with the server URL (https://mcp.contentin.io/mcp-server) and a ContentIn sign-in. Gemini connects through Gemini CLI. OpenClaw and n8n connect with the same URL and an API key in the X-MCP-Key header. It also works in Cursor and VS Code. Any other agent that speaks MCP over Streamable HTTP can use the same URL and key.
How do I use it in n8n?: Create an API key in ContentIn under Profiles, Connections, API keys (Growth or Pro). In n8n, add an MCP Client node, set the transport to HTTP Streamable, use the server URL as the endpoint and add a Header Auth credential with the name X-MCP-Key and your key as the value. The tools then show up for your workflow or AI Agent node. Each key acts for one ContentIn profile.
Can an automation publish without me?: No. Scheduling and publishing are two-step: the tool returns the exact text and a confirmation token, and the post only goes out once a person has confirmed that text. Automations are good for gathering analytics and leads, saving material and producing drafts; the last step stays with you.
Is it LinkedIn-safe? Does it use the official API?: Yes. Posting and post analytics go through LinkedIn's official API for the LinkedIn account you connected to ContentIn. Your agent never logs into your LinkedIn account, and your LinkedIn password and cookies never reach ContentIn or your agent, so there is none of the session automation that gets accounts restricted.
Do I need to write prompts or configure the tools?: No. Once connected, you ask in plain language and your agent picks the right tool. There's no system prompt to write. If you build your own agent or workflow, the tool descriptions come with the server.
I run several profiles or an agency. How does that work?: Each connection or API key is bound to one LinkedIn profile, which you pick when you connect. Agencies connect one per client profile, so each client's voice, material and queue stay separate.
Are the Coming next items available?: No. Post library search, draft check, webhooks and the Agent + Leads plan are not available yet. We build them in the order people ask for them; the “I want this” buttons on this page are how we count that.
What happens to my data?: A connected agent can only reach the one ContentIn profile you approved: its posts, drafts, ideas, analytics, comments and leads. It can't see your other profiles, your billing details or your LinkedIn password. You can see and revoke every connection and API key in ContentIn under Profiles, Connections, and revoking takes effect immediately. What your agent does with data it reads is governed by that agent's own terms.

Keep your agent. Make it write like you.

Drop your LinkedIn profile, let ContentIn learn how you write, then connect your agent or workflow with one URL.