Why am I seeing "Agent initialization failed" or "Blob storage initialization error"?
Why am I seeing "Agent initialization failed" or "Blob storage initialization error"?
Example errors:If you installed via the curl script, this error should not occur because the binary bundles its own runtime. Re-run the install script to fix:If you installed via npm, this is caused by a Node.js version mismatch — the CLI was installed with a different Node version than what’s currently active.Solution: Use Node.js >= 20 and reinstall:After reinstalling, run
From version 2.3.2 onwards, the SQLite error above no longer occurs — the
better-sqlite3 dependency was removed. If you see this error, updating to the latest version will resolve it.brv restart to start fresh.Why does the TUI show "running:0" when a task is actually running?
Why does the TUI show "running:0" when a task is actually running?
Symptom:A command (like Alternatively, press
/query) is visibly running in the REPL, but the TUI displays running:0.Cause:This typically happens when:- The REPL has been kept active for a long time
- Your laptop went to sleep mode while the REPL was open
brv restart to stop all processes and start fresh:Ctrl+C to quit the current session and start a new one with brv.I'm getting errors when using a third-party provider model. How do I fix it?
I'm getting errors when using a third-party provider model. How do I fix it?
Third-party providers may return errors due to API key issues, unsupported features, rate limits, or model changes. The quickest fix is to switch back to ByteRover’s built-in LLM — it’s fully compatible with all ByteRover features. It requires a logged-in ByteRover account (no API key):Or in the REPL, run Or in the REPL, run
“Unsupported feature” or tool-use errorsNot all models support tool use (function calling). If you see errors related to tool calls or unsupported features, switch to a model from the recommended list — these have been verified to work with ByteRover’s full feature set.
Provider API is down or returning 500 errorsThis is on the provider’s side. Check their status page (e.g., status.openai.com, status.anthropic.com) and switch to another provider in the meantime:
Streaming errors show
/providers and select ByteRover (you’ll be prompted to sign in if not already logged in).If you prefer to keep using an external provider, make sure to pick a model from the recommended LLM list — these have been verified for full compatibility. To switch models:/model to browse and select.Below are specific errors and their workarounds.“Unsupported feature” or tool-use errorsNot all models support tool use (function calling). If you see errors related to tool calls or unsupported features, switch to a model from the recommended list — these have been verified to work with ByteRover’s full feature set.
Provider API is down or returning 500 errorsThis is on the provider’s side. Check their status page (e.g., status.openai.com, status.anthropic.com) and switch to another provider in the meantime:
Streaming errors show
[object Object] or unclear messagesIf you see [object Object] instead of a meaningful error message during streaming, update to v2.3.3 or later. This version improves error handling for provider streaming failures — particularly for models like GPT-5. While it may not catch every case, most streaming errors should now display a more useful message that helps you diagnose the issue.brv providers connect or brv providers switch shows success but provider doesn’t workPrior to v2.5.1, these commands could report success even when the server rejected the request (e.g., due to missing authentication). Update to v2.5.1 or later — authentication and server-side errors are now correctly surfaced in the command output.How do I configure ByteRover CLI to work behind a corporate proxy?
How do I configure ByteRover CLI to work behind a corporate proxy?
Enterprise proxy support requires ByteRover CLI v2.4.0 or later.
brv:| Variable | Purpose | Example |
|---|---|---|
HTTPS_PROXY | Proxy for HTTPS traffic | http://proxy.corp.com:8080 |
HTTP_PROXY | Proxy for HTTP traffic | http://proxy.corp.com:8080 |
NO_PROXY | Comma-separated list of hosts that bypass the proxy | localhost,127.0.0.1,.internal.corp.com |
NODE_EXTRA_CA_CERTS | Path to corporate CA certificate for SSL inspection | /etc/ssl/certs/corp-ca.pem |
Common proxy errors and fixes:
| Symptom | Likely cause | Fix |
|---|---|---|
UNABLE_TO_VERIFY_LEAF_SIGNATURE, SELF_SIGNED_CERT_IN_CHAIN, or CERT_HAS_EXPIRED | Corporate SSL inspection without a trusted CA cert | Set NODE_EXTRA_CA_CERTS to your corporate CA certificate path |
Connection timeouts (ECONNABORTED, ETIMEDOUT, ECONNREFUSED) | Proxy not set or incorrect proxy URL | Verify HTTPS_PROXY is set and the proxy host/port are correct |
407 Proxy Authentication Required | Proxy requires credentials | Include credentials in the URL: http://user:[email protected]:8080 |
Proxy routing only applies to ByteRover’s internal services — cloud sync, hub, and authentication. Traffic to external LLM providers (OpenAI, Anthropic, OpenRouter, etc.) is not routed through the proxy.Commands like
brv query and brv curate route through the proxy when using the ByteRover built-in provider. To set this up, run brv login followed by brv providers connect byterover.How do I diagnose issues with ByteRover?
How do I diagnose issues with ByteRover?
Run Common status messages and fixes:
If
brv status for a quick health check. It shows your CLI version, account state, current space, and context tree status:| Message | Fix |
|---|---|
| ”Daemon failed to start automatically” | Run brv restart |
| ”Daemon crashed unexpectedly” | Run brv restart |
| ”Account: Not connected” | Normal for local-only usage — no action needed |
brv restart does not resolve the issue, try reinstalling: