Skip to content

Troubleshooting

Solutions to common issues in HiveMind OS. If your problem isn't listed here, check the FAQ or file a bug report.

1. Daemon Not Starting

Symptoms: CLI commands fail with "connection refused" or "daemon not running."

Fixes:

  • Check the system tray — the HiveMind OS icon should be visible. Click it and select Start Daemon.
  • Restart manually:
    bash
    hive daemon stop
    hive daemon start
  • Check logs for errors:
    bash
    cat ~/.hivemind/logs/daemon.log | tail -50
  • Ensure port 9180 is not in use by another process.

WARNING

On Linux, if the daemon crashes at startup, verify that your DISPLAY or WAYLAND_DISPLAY environment variable is set correctly.

2. "No Secret Found" After Setting Up a Provider

Symptoms: You added a provider API key in Settings → Providers, but using it gives:

no secret found in OS keyring for key `provider:...:api-key`

Fixes:

  • Save your config after adding the key. The daemon reloads secrets when the configuration is saved. Click Save in Settings after entering your API key.
  • If you just added the key, try switching to a chat session and sending a message — the config save triggers a reload automatically.
  • As a last resort, restart the daemon from the system tray or via hive daemon stop && hive daemon start.

3. Connector "No OAuth Tokens" After Authorization

Symptoms: You authorized a connector (e.g., Microsoft 365) successfully and the test passed, but the agent reports:

no OAuth tokens for connector 'microsoft'. Please re-authorize via Settings → Connectors.

Fixes:

  • Re-authorize from Settings → Connectors — click the Authorize button again.
  • Ensure you're running the latest version. Earlier releases had a bug where secrets could be lost when saving other settings.
  • Restart the daemon: hive daemon stop && hive daemon start, then re-authorize.

4. Provider Connection Failed

Symptoms: "Provider unreachable" or "401 Unauthorized" errors.

Fixes:

  • Verify your API key is set correctly in Settings → Providers.
  • Check that environment variables are loaded (if using env:VAR_NAME in provider auth config).
  • For Ollama, ensure the server is running: ollama serve.

5. MCP Server Not Connecting

Symptoms: Tools from an MCP server are unavailable or show "disconnected."

Fixes:

  • Verify the command path exists and is executable:
    bash
    which npx  # or the command in your config
  • Check the MCP server config in your persona's mcp_servers section (visible in Settings → Personas).
  • Restart the connection from the persona's MCP Servers section in Settings → Personas.
  • Check MCP logs:
    bash
    cat ~/.hivemind/logs/mcp-*.log | tail -30

6. Agent Seems Stuck

Symptoms: The agent shows a spinner but produces no output for a long time.

Fixes:

  • Check the tool approval queue — the agent may be waiting for you to approve a tool call. Look for a pending approval badge in the UI.
  • Press Ctrl+C to stop the current agent and retry.
  • If the agent loops, try simplifying your request or switching to a different model.

TIP

Configure security.default_permissions in your config to pre-approve trusted tools and avoid blocking on approvals. Permission rules specify tool patterns with auto, ask, or deny actions.

7. Knowledge Not Being Recalled

Symptoms: Asking the agent about previously saved information returns no results.

Fixes:

  • Try broader or different search terms when asking the agent about stored knowledge.
  • Use the Knowledge Explorer in the UI to browse and search saved knowledge.
  • Check that the knowledge entry was saved — open the Knowledge panel and search for it.

8. High Memory Usage

Symptoms: HiveMind OS uses excessive RAM, causing system slowdowns.

Fixes:

  • Use smaller local models (e.g., llama3:8b instead of llama3:70b).
  • Close unused chat sessions — each session holds context in memory.
  • Restart the daemon to clear accumulated state:
    bash
    hive daemon stop && hive daemon start

9. Bot Not Responding

Symptoms: A launched bot doesn't reply to messages or seems idle.

Fixes:

  • Check bot status in the Bots page in the UI.
  • Verify the persona configuration is valid — a missing or broken system prompt can cause silent failures.
  • Check that the persona's assigned provider is connected in Settings → Providers.
  • Stop and relaunch the bot from the Bots page.

10. Workflow Not Triggering

Symptoms: A workflow that should run automatically never starts.

Fixes:

  • Check the workflow configuration in the Workflows page in the UI (workflows are managed through the UI/API, not config files).
  • For event-based triggers, verify the event source is connected and emitting events.
  • Try running the workflow manually from the Workflows page in the UI.
  • Check workflow run history on the Workflows page for errors.

TIP

Run hive daemon status to confirm the daemon is running and healthy.

See Also

Released under the MIT License.