Model Context Protocol

LinkedIn analytics for your AI agent

Blabigo exposes an MCP server so Claude and other AI agents can read your LinkedIn analytics on your behalf — securely, with your own credentials, no scraping.

Get your API key free →What is MCP?

Server endpoint

https://studio.blabigo.ai/api/mcp/linkedin

JSON-RPC 2.0 over HTTP (protocol version 2024-11-05). Standards-based OAuth discovery is published at /.well-known/oauth-protected-resource, so compliant clients can connect automatically.

Available tools

linkedin_list_accounts

List the user's active LinkedIn accounts and managed company Pages.

linkedin_get_post_analytics

Post-level analytics — impressions, clicks, engagement — with optional date and account filters.

linkedin_get_account_summary

Per-account analytics summary aggregated over a date range.

linkedin_refresh_analytics

Pull the latest metrics from LinkedIn for the user's published posts.

linkedin_list_posts

List posts — drafts, scheduled, published, failed — with status, date and origin filters.

linkedin_get_post

Get one post by id, including its schedule and publishing state.

linkedin_create_draft

Create a post as a draft. Drafts never publish on their own.

linkedin_schedule_post

Queue a post for a future time. Blabigo's scheduler publishes it then — this does not publish immediately.

linkedin_reschedule_post

Move an already-scheduled post to a different time.

linkedin_cancel_scheduled_post

Withdraw a scheduled post. It reverts to a draft — nothing is deleted.

The analytics tools are read-only; linkedin_refresh_analytics pulls fresh metrics from LinkedIn. Tools always resolve data for the authenticated user — an agent can never read another account.

What an agent cannot do

Handing an agent your LinkedIn account is only reasonable if the blast radius is bounded. Here is exactly where the limits are.

It cannot publish instantly

Every post an agent creates lands on your queue with a scheduled time. Blabigo's scheduler publishes it at that time. No tool reaches LinkedIn directly, so there is always a window in which you can look at what is queued.

It cannot delete anything

Cancelling a scheduled post reverts it to a draft. The post stays in your dashboard, so you can see exactly what an agent withdrew and why. There is no delete tool and no delete endpoint.

It cannot exceed your quotas

Per-identity daily caps, a minimum gap between posts, and a scheduling horizon are enforced server-side. They bind agents, the API, the CLI and the web app equally, so a runaway loop hits a wall rather than your followers.

It cannot hide

Agent-created posts carry a badge naming the credential that made them. Settings → MCP shows what each key has scheduled, and lets you pull an entire key's queue back to drafts in one click.

It cannot exceed its grant

OAuth clients get scoped tokens. A client granted only analytics access cannot see the content tools at all, let alone call them. There is no publish scope to grant, because there is nothing to grant it to.

It cannot reach another account

Every tool resolves data for the authenticated user and the identities they have been granted access to. Permissions are re-checked at publish time, not just when the post is queued.

Authentication

API key

Generate a key in Settings → MCP and send it as a Bearer token. Best for scripts and self-hosted agents.

OAuth 2.0 + PKCE

Register a client and let end users authorize with the standard authorization-code + PKCE (S256) flow. Best for hosted AI apps.

Connect

Option A — paste the URL (most clients)

In Claude, ChatGPT and other connector-based clients, add a custom connector pointing at the endpoint above. The client signs you in and asks your permission — there is no key to create or store.

Claude Code

claude mcp add --transport http blabigo https://studio.blabigo.ai/api/mcp/linkedin

Config-file clients

{
  "mcpServers": {
    "blabigo": {
      "url": "https://studio.blabigo.ai/api/mcp/linkedin"
    }
  }
}

Option B — API key, for clients without remote or OAuth support

A key never expires and carries full access. Prefer Option A unless your client cannot do either.

1. Create a key

In the dashboard, go to Settings → MCP access and create a key. It starts with mcp_live_ and is shown once.

2. Store it on your machine

npx -y @blabigo/cli auth login --key mcp_live_your_key_here

Written to ~/.config/blabigo/config.json with owner-only permissions. In CI, set BLABIGO_API_KEY instead.

3. Add it to your MCP client config

{
  "mcpServers": {
    "blabigo": {
      "command": "npx",
      "args": ["-y", "@blabigo/cli", "mcp"]
    }
  }
}

Works in any MCP client that can launch a command — Claude Desktop, Claude Code, and others. The command speaks MCP on stdio and forwards to https://studio.blabigo.ai/api/mcp/linkedin with your key.

Or call the HTTP API directly

curl -X POST https://studio.blabigo.ai/api/mcp/linkedin \
  -H "Authorization: Bearer mcp_live_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

What an agent cannot do

No tool publishes to LinkedIn immediately, and no tool deletes a post. Agents queue; Blabigo’s scheduler publishes at the scheduled time. Cancelling reverts a post to a draft, so you can always see what an agent pulled back.

Ready to plug LinkedIn analytics into your AI workflow?

Create a free account →

More free tools

LinkedIn Post Preview

See exactly how your post will look in the LinkedIn feed — with the “…more” cut-off, hashtags and mentions styled — before you hit publish.

Blabigo is an independent product. It is not affiliated with, authorized by, endorsed by, or sponsored by LinkedIn Corporation or Microsoft. “LinkedIn” is a registered trademark of LinkedIn Corporation, used here only descriptively to identify the third-party service this software works with.