Connect the LinkedIn API
Integrate the LinkedIn API through Modiva’s unified REST API and hosted MCP server. Share posts and read company-page analytics. OAuth, token refresh, rate limits and 15 typed tools are handled for you — connect once and call LinkedIn from any agent or app through a single endpoint.
What you can do with the LinkedIn API
- ✓ Publishing
- ✓ Scheduling
- ✓ Inbox & DMs
- ✓ Analytics
- ✓ Webhooks
LinkedIn MCP tools & scopes
- accounts_
- list · health · status
- posts_
- create · cross_post · schedule · publish_now
- comments_
- list · reply
- messages_
- list · send · inbox
- analytics_
- unified · best_time_to_post
OAuth scopes requested
- Read profile & accounts
- Publish posts & stories
- Read insights
- Read & reply to comments/DMs
Connect LinkedIn in your agent
# Add Modiva's MCP server to your agent
{
"mcpServers": {
"modiva": {
"url": "https://mcp.modiva.ai",
"headers": { "Authorization": "Bearer $MODIVA_MCP_KEY" }
}
}
}
# Then connect LinkedIn and call its tools, e.g.
accounts_list(provider="linkedin")Frequently asked questions
How do I connect the LinkedIn API with Modiva?
Create a Modiva account, generate an MCP key, and connect LinkedIn from your dashboard. Modiva runs the full OAuth flow and stores the encrypted token for you — no client secrets in your code.
What can I do with the LinkedIn API through Modiva?
Share posts and read company-page analytics. You get 15 typed tools grouped by resource (accounts_, posts_, comments_, messages_, analytics_), callable over REST or the MCP server.
Do I need to handle OAuth or token refresh myself?
No. Modiva handles the LinkedIn OAuth handshake, token storage (envelope-encrypted) and automatic refresh. You call a single unified endpoint.
Can an AI agent use the LinkedIn API?
Yes — that is the point. Modiva exposes LinkedIn as Model Context Protocol (MCP) tools, so agents like Claude, Cursor or your own app can read and act on LinkedIn through one endpoint.
LinkedIn playbooks & guides
Free, step-by-step playbooks on growing with LinkedIn — what works, plus how to run it with Modiva.