Reeve
Help

Troubleshooting

Fix common Reeve issues — connection problems, auth errors, connector sync failures, and more.

Troubleshooting

Quick fixes for the most common issues. If yours isn't listed here, check the FAQ or ask on Discord.

Start Here

Run these to get a snapshot of what's wrong:

reeve status          # Quick health check
reeve status --all    # Detailed report (safe to share — tokens are redacted)
reeve doctor          # Auto-diagnose and repair common issues

reeve status --all is the best thing to paste when asking for help. All sensitive values are redacted automatically.


Connection Issues

Desktop App shows "Connecting…" or "Disconnected"

  1. Quit the app fully and relaunch
  2. Run reeve gateway status in a terminal to check if the gateway process is running
  3. Check for port conflicts: another service may be using port 18789
  4. Run reeve doctor — it detects and repairs common config issues

Cockpit shows "Unauthorized"

The Cockpit token doesn't match the gateway token.

  1. Run reeve cockpit — this opens the Cockpit with the correct token embedded in the URL
  2. If using a remote gateway, confirm gateway.auth.token matches your connect config
  3. Clear browser storage for app.meetreeve.com and reload

Can't reach the Cockpit remotely

  • Recommended: Use Tailscale Serve — reeve gateway --tailscale serve
  • SSH tunnel: ssh -N -L 18789:127.0.0.1:18789 user@your-server, then open http://127.0.0.1:18789
  • Check that gateway.bind is set correctly for your access method

Authentication & Sign-In

Agent gives no reply / "all models failed"

  1. Run reeve models status — checks LLM provider connectivity
  2. Verify API keys are configured: reeve config get models.providers
  3. Check your Anthropic or OpenAI account — expired billing or rate limits cause silent failures
  4. Add a fallback model so Reeve can switch providers when one is unavailable
  1. Check your spam/junk folder
  2. Try signing in with Google instead
  3. Clear browser cookies for meetreeve.com and retry

Rate limit errors (HTTP 429)

You've hit the API provider's rate limit.

  1. Wait 1–5 minutes for the window to reset
  2. Configure a fallback model from a different provider
  3. Upgrade your API plan for higher rate limits at the provider (Anthropic Console or OpenAI Platform)

Connector Issues

A connector shows "Disconnected"

  1. Go to Connectors in the Cockpit and click Reconnect
  2. For OAuth connectors (Shopify, Meta Ads, Google Ads): the token may have expired — re-authorize
  3. For API key connectors (Klaviyo, Zendesk, Gorgias): verify the key hasn't been rotated or deleted
  4. Check that the external platform hasn't suspended or restricted your account

Shopify data is missing or stale

  1. Confirm your store domain uses the *.myshopify.com format, not a custom domain
  2. Check that the Reeve app is still installed in your Shopify admin (Settings → Apps)
  3. Disconnect and reconnect from the Connectors page
  4. Shopify has a short API cache — very recent changes may take a few minutes to appear

Ad data is stale or missing

Ad platforms typically have a 1–4 hour reporting delay — this is normal and not a Reeve issue.

  1. Ask Reeve for a fresh report to trigger a new API fetch
  2. Verify your Meta or Google ad account is active and not restricted
  3. Check connector status in the Cockpit — reconnect if needed

Klaviyo emails / metrics not showing

  1. Verify the API key has read access to campaigns, flows, and metrics
  2. Disconnect and reconnect the Klaviyo connector
  3. Confirm the API key is a private key (not a public key)

Agent & Chat Issues

Sent a message, got no reply

  1. reeve status — is the gateway up?
  2. reeve models status — can the agent reach an LLM?
  3. If using WhatsApp or Telegram: check if a pairing approval is pending (reeve pairing list)
  4. reeve logs --follow — look for errors in the live log stream

"Context too large" errors

The conversation has filled the model's context window.

  1. Start a fresh session: send /new in the chat
  2. Reeve auto-compacts context when it gets close to the limit — this is normal behavior
  3. If it happens repeatedly, see Compaction

Slow responses

  1. Check your LLM provider's status page — degraded API performance is usually the cause
  2. Try a faster model: reeve config set agents.defaults.model "claude-sonnet-4-5"
  3. Verify your internet connection is stable

Desktop App Issues

App won't open ("unidentified developer")

Right-click the app icon → OpenOpen again. macOS Gatekeeper only shows this once — the app is signed and notarized.

Blank white window on launch

  1. Quit the app fully (⌘Q, not just close window)
  2. Delete ~/Library/Application Support/Reeve and relaunch
  3. Check Console.app for crash logs mentioning "Reeve"

Updates not installing

  1. The app checks for updates on launch — quit and relaunch to trigger a check
  2. Download the latest version manually from meetreeve.com

Still Stuck?

  1. Run reeve status --all and copy the output
  2. Discord: discord.gg/meetreeve — fastest community response
  3. GitHub: github.com/MindFortressInc/reeve/issues — for confirmed bugs
  4. Email: support@meetreeve.com — for account-specific issues

On this page