Could not add provider"- no backend logs, all scopes and config verified

Description

When trying to connect a LinkedIn Company Page channel, the OAuth flow completes successfully (LinkedIn consent screen appears with all correct scopes, user clicks Allow), but Postiz redirects back with "Could not add provider" error. No details are logged in the backend.

Environment

• Postiz version: latest (ghcr.io/gitroomhq/postiz-app:latest, pulled March 17, 2026)
• Deployment: Docker Compose on Contabo VPS (Ubuntu)
• Reverse proxy: Nginx (built into Postiz container)
• SSL: Cloudflare tunnel

Configuration

• LINKEDIN_CLIENT_ID and LINKEDIN_CLIENT_SECRET set in docker-compose.yml environment
• LinkedIn Developer App: "KMDS Postiz", verified with Company Page
• Products enabled: Share on LinkedIn, Sign In with LinkedIn using OpenID Connect, Advertising API
• All OAuth 2.0 scopes present: rw_organization_admin, w_member_social, r_profile_basicinfo, rw_events, r_ads, r_basicprofile, r_organization_admin, email, r_1st_connections_size, openid, profile, r_ads_reporting, r_organization_social, r_verify, w_organization_social, rw_ads, r_events
• Redirect URL configured: https://postiz.km-digitalsolutions.com/integrations/social/linkedin-page
• User is ADMINISTRATOR on the LinkedIn Company Page

OAuth URL from browser (decoded)

The LinkedIn authorization URL correctly contains: Steps to reproduce

1. Go to Postiz → Channels → Add Channel → LinkedIn Page
2. Redirected to LinkedIn login
3. Sign in with admin account for Company Page
4. LinkedIn consent screen shows correct permissions (organization posting, profile access, etc.)
5. Click "Allow"
6. Redirected back to Postiz
7. Error: "Could not add provider"

What I've checked

docker logs postiz --since 30s 2>&1 shows NO LinkedIn-related entries after the callback — only OpenAI API key warnings docker logs postiz 2>&1 | grep -i "linkedin|scope|NotEnough|error|403|provider" returns only startup logs, no callback errors Tried both "Not enough scopes" (first attempts before update) and "Could not add provider" (after pulling latest image) Cleared browser cache, tried incognito, hard refreshed Restarted all containers multiple times Pulled latest image on March 17, 2026

Possibly related issues

#972 (LinkedIn 403 profile picture fetch — may still be the root cause, error just isn't logged) #1197 (NotEnoughScopes for personal profile — different provider but similar symptoms) #1134 (PR that decoupled personal/page scopes — may not have fully resolved page provider)

What I suspect

The backend is failing silently during the OAuth callback. The error likely occurs in the authenticate() method of linkedin.page.provider.ts but is caught and swallowed without logging. This could be the profile picture 403 issue from #972 which may not be fully fixed in the latest image.

Request

Could the team add error logging to the LinkedIn Page provider's authenticate/callback flow so self-hosters can diagnose these failures? Currently the "Could not add provider" message gives no actionable information.