linkedin-scheduler-remote-ldraney 0.2.0

Remote MCP connector wrapping linkedin-mcp-scheduler with LinkedIn OAuth 2.0 over Streamable HTTP

linkedin-scheduler-remote

Remote MCP server wrapping linkedin-mcp-scheduler with LinkedIn OAuth 2.0 and Streamable HTTP transport — designed for Claude.ai connectors.

How it works

Claude.ai ──HTTP+Bearer──> linkedin-scheduler-remote ──LinkedIn API──> LinkedIn
                                      │
                                imports linkedin-mcp-scheduler's FastMCP (all 8 tools)
                                patches get_client() with per-request ContextVar
                                adds LinkedIn OAuth + /health + /oauth/callback

Three-party OAuth: Claude <-> this server <-> LinkedIn. Each user authorizes their own LinkedIn account via OAuth. The server stores per-user LinkedIn access tokens (encrypted at rest).

Prerequisites

1. A LinkedIn Developer App at LinkedIn Developer Portal:
   • Create an app (or use existing)
   • Under Auth settings, add authorized redirect URL: {BASE_URL}/oauth/callback
   • Note the Client ID and Client Secret
   • Ensure the app has the products: "Share on LinkedIn" and "Sign In with LinkedIn using OpenID Connect"
2. Scopes required: openid, profile, email, w_member_social

Setup

git clone https://github.com/ldraney/linkedin-scheduler-remote.git
cd linkedin-scheduler-remote
cp .env.example .env
# Edit .env with your LinkedIn OAuth credentials, BASE_URL, and SESSION_SECRET
make install
make run

Environment Variables

Variable	Description
LINKEDIN_OAUTH_CLIENT_ID	LinkedIn app Client ID
LINKEDIN_OAUTH_CLIENT_SECRET	LinkedIn app Client Secret
SESSION_SECRET	Random secret for encrypting token store
BASE_URL	Public HTTPS URL where this server is reachable
HOST	Bind address (default: 127.0.0.1)
PORT	Listen port (default: 8002)

Verification

curl http://127.0.0.1:8002/health                                    # {"status": "ok"}
curl http://127.0.0.1:8002/.well-known/oauth-authorization-server    # OAuth metadata
curl -X POST http://127.0.0.1:8002/mcp                               # 401 (auth required)

Tools (8)

All scheduling tools from linkedin-mcp-scheduler are exposed:

• schedule_post — Schedule a LinkedIn post for future publication
• list_scheduled_posts — List posts with optional status filter
• get_scheduled_post — Get details of a scheduled post
• cancel_scheduled_post — Cancel a pending post
• update_scheduled_post — Edit pending post fields
• reschedule_post — Change the scheduled time
• retry_failed_post — Reset failed posts to pending
• queue_summary — Get queue statistics

Deploying

Standalone

Use any HTTPS tunnel (Tailscale Funnel, ngrok, Cloudflare Tunnel) to expose the server publicly, then add the URL as a Claude.ai connector.

A systemd service file is provided in systemd/linkedin-scheduler-mcp-remote.service.

Kubernetes (production)

Production deployment is managed by mcp-gateway-k8s, which runs this server as a pod with Tailscale Funnel ingress.

Design Notes

The scheduler-remote differs from gcal/notion remotes in one key way: there's a daemon (linkedin-mcp-scheduler-daemon) that publishes posts to LinkedIn. Currently operates in single-user mode — the daemon uses its own env-var credentials. Multi-user daemon support (per-user credentials stored with each post) is a future enhancement.

Privacy

See PRIVACY.md for our privacy policy.

License

MIT