Research
Find patterns in posts that perform
Ask Claude to search Postiv's inspiration library, group winning hooks and formats, and turn the patterns into original ideas.
Hosted LinkedIn MCP, live now
LinkedIn MCP Server for Claude and AI Agents
A LinkedIn MCP server is a secure connector that gives Claude and other AI agents a defined set of LinkedIn capabilities. Postiv lets your agent research, create, design, schedule, and analyze content without controlling linkedin.com through a browser.
Works with
Claude Code
Claude
Codex
ChatGPT
Cursor
Windsurf
VS Code
OpenAI Agents
Vercel AI SDK
LangChain
n8n
Make
Zapier
Claude Code - postiv linkedin mcp
$ claude
4 of 25 MCP tools shown / 1 LinkedIn content workspace
search_knowledge
Pulled positioning, past posts, and brand rules from the workspace.
create_post
Draft ready, grounded in real workspace context.
create_carousel
Narrative, slide structure, and copy ready to render.
schedule_post
Queued for publishing and routed to any configured approvers.
Ready for MCP clients. No prompt glue. No separate LinkedIn tools.
Real setup and workflow demo
See Claude use LinkedIn tools through Postiv
Watch the complete connection process, then see Claude research successful posts, create content, and analyze LinkedIn performance inside one conversation.
Create
Turn a conversation into a real draft
Create a post in the correct Postiv workspace, refine it in Claude, then open the saved draft for visual review.
Analyze
Explain performance in plain English
Pull cached LinkedIn metrics into Claude, compare periods, and identify the content experiments worth testing next.
Choose the right connector
A content workflow, not a browser scraper
Many tools use the phrase LinkedIn MCP for different jobs. Postiv is designed for content teams that want controlled research, creation, visuals, scheduling, and analytics. It does not automate DMs or connection requests.
| Capability | Postiv MCP | Scraper MCP | Analytics connector |
|---|---|---|---|
| Research content inspiration | Yes | Varies | Usually no |
| Create grounded LinkedIn posts | Yes | Varies | Usually no |
| Generate images and carousels | Yes | Usually no | No |
| Schedule and reschedule posts | Yes | Varies | Usually no |
| Respect human approval rules | Yes | Varies | No |
| Analyze content performance | Yes | Varies | Yes |
| Automate DMs or connection requests | No | Often | No |
Connect in three steps
Connect Claude to LinkedIn with Postiv
The hosted connector requires no local server. Follow the visual tutorial for the complete walkthrough or use the technical reference for exact client configuration.
Connect LinkedIn to Postiv
Choose the personal profile or Company Page that Claude should work with. Postiv remains the controlled content layer.
Generate a Postiv API key
Open Settings, then Integrations, and create an organization key. Treat the key like a password.
Add the hosted MCP to Claude
Add Postiv as a custom connector, start a new chat, and test the connection with a read-only request first.
Step 0. Get your API key
In Postiv, go to Settings, then Integrations, then the Public API section, and generate a key. The same key works for the MCP server and the REST API. Only admins can generate or view it.
Get your API key
Add Postiv as an MCP in ChatGPT
ChatGPT connects to Postiv with the hosted MCP URL and an Authorization header. Treat the header value like a password.
- 1. In ChatGPT, open Settings, then Plugins, then MCPs.
- 2. Add an MCP server, name it Postiv, and enter https://postiv.ai/mcp as the URL.
- 3. Add an Authorization header with the value Bearer followed by your Postiv API key, then save.
URL
https://postiv.ai/mcpHeader
Authorization: Bearer pk_postiv_YOUR_KEYLeave the Bearer token environment-variable field empty unless you have configured a local environment variable for the key.
Full endpoint reference, scopes, and the REST API live in the API documentation.
Copy and try
Six useful Claude prompts for LinkedIn
Start with a read-only research request. Move to creation and scheduling only after you have confirmed the correct Postiv workspace and LinkedIn profile.
ResearchPrompt 1
Find the strongest educational posts in my inspiration library from the last 90 days. Group them by hook type and format, then explain the patterns without copying the wording.
AnalyticsPrompt 2
Show my top five LinkedIn posts from the last 90 days and explain the topics, structures, and calls to action they have in common.
WritingPrompt 3
Draft a LinkedIn post about [topic] using my workspace knowledge and normal voice. Ask me two clarifying questions before creating the draft.
RepurposingPrompt 4
Take my best-performing post from the last 90 days and propose three new angles. Keep the core insight, but do not reuse the original wording.
CarouselPrompt 5
Turn the approved post into a five-slide LinkedIn carousel outline. Show me the slide copy first and wait for approval before rendering it.
SchedulingPrompt 6
List my approved, unscheduled drafts and recommend a balanced publishing order for next week. Do not schedule anything until I confirm the plan.
Inside the MCP server
Postiv workflows become agent-callable tools
Your agent can create useful LinkedIn work with real workspace context, explicit actions, and one connected content system.
postiv-linkedin-mcp
25 toolsget_workspace
search_knowledge
search_inspiration
create_post
schedule_post
create_carousel
generate_post_image
get_analytics_summary
Workspace context
Confirm the connected workspace and available scopes before an agent takes action.
Knowledge grounding
Save and retrieve real positioning, product facts, and brand knowledge instead of asking an agent to guess.
Inspiration research
Search a moderated library of high-performing LinkedIn content before creating an original draft.
Profiles and Comment Plug
Choose the correct LinkedIn profile or Company Page and manage its automatic first comment.
Post drafting and editing
Create, list, retrieve, and update saved LinkedIn posts through a controlled draft lifecycle.
Scheduling with approvals
Schedule, reschedule, and unschedule posts while respecting approvers configured on the target profile.
AI image generation
Create a branded image for a post, then iterate on it with plain-language instructions.
Carousel generation
Turn a structured outline into a branded LinkedIn carousel and edit individual slides before publishing.
Performance analytics
Read cached LinkedIn performance summaries, top posts, account trends, and post-level analytics.
Interactive widgets
Supported MCP hosts can render Postiv post editors, image previews, carousel pagers, and inspiration cards inline.
LinkedIn-safe formatting
Postiv converts formatting into LinkedIn-safe text so published posts display as intended.
Controlled publishing
Scheduling always queues the post, and configured approval rules still apply before it can go live.
Security and control
Keep credentials private and humans in control
The MCP makes capabilities available to your agent. Your Postiv workspace, key scopes, target profile, and approval rules determine what happens next.
Scoped workspace access
The API key controls which organization and capabilities the agent can access.
Review before scheduling
Ask Claude to show the final copy, target profile, date, and time before it queues a post.
Approval rules stay active
Posts still wait for approval when the selected LinkedIn profile has approvers configured.
Rotate exposed credentials
Never share the key or connector URL. Regenerate the key immediately if either is exposed.
LinkedIn MCP FAQ
Questions about connecting Claude to LinkedIn
What is a LinkedIn MCP server?
A LinkedIn MCP server is a connector that gives an AI assistant a defined set of LinkedIn-related capabilities. Postiv's hosted server focuses on content workflows such as research, drafting, visuals, scheduling, approvals, and analytics.
Can Claude connect to LinkedIn?
Yes. Claude can connect to Postiv's hosted LinkedIn MCP server. Postiv then gives Claude controlled access to the LinkedIn content workflows and connected profiles available in your Postiv workspace.
What can Claude do after connecting Postiv?
Claude can research inspiration, use workspace knowledge, create and edit posts, generate images and carousels, schedule approved content, and analyze cached LinkedIn performance data.
Can Claude publish a LinkedIn post without my approval?
Postiv's schedule action queues a post instead of publishing it instantly. If the target profile has approvers configured, its existing approval workflow still applies. You can also instruct Claude to show the final post before scheduling.
Does Postiv scrape LinkedIn or automate DMs?
No. Postiv's LinkedIn MCP is built for content research, creation, scheduling, and performance workflows. It is not designed for profile scraping, connection requests, or automated direct messages.
Does the LinkedIn MCP work with Claude Code?
Yes. Claude Code can connect to the hosted MCP endpoint with an Authorization header. The exact command and authentication options are available in the hosted MCP documentation on Postiv.
Can Claude work with a LinkedIn Company Page?
Yes, provided the Company Page is connected to your Postiv workspace and your account has permission to manage it. Confirm the target profile before creating or scheduling content.
How should I protect my Postiv API key?
Treat the key and any connector URL containing it like a password. Never put either in a public screenshot, post, transcript, or support message. Regenerate the key immediately if it is exposed.
Give Claude controlled LinkedIn capabilities
Connect your Postiv workspace, verify the right profile, and turn your next LinkedIn idea into an approved content workflow.