API Errors

Handling errors from AI providers.

Rate Limiting

"rate_limit_exceeded"

Cause: Too many requests.

Solution:

• Wait 60 seconds
• Upgrade your plan
• Reduce request frequency

"rate_limit_request_enqueued"

Cause: Request queued due to high load.

Solution:

• Wait for automatic retry
• Reduce concurrent requests

Authentication

"invalid_api_key"

Cause: Wrong API key.

Solution:

• Check key format (sk-ant-...)
• Generate new key
• Verify key in settings

"permission_denied"

Cause: Key lacks required permissions.

Solution:

• Check key scope
• Generate key with correct permissions
• Contact Anthropic support

Content Errors

"context_length_exceeded"

Cause: Too much context.

Solution:

> /compact
> Let's continue with the compacted context

"invalid_request_error"

Cause: Malformed request.

Solution:

• Update to latest version
• Check settings.json format
• Report bug if persistent

Server Errors

"api_error" or "server_error"

Cause: Provider issue.

Solution:

• Wait and retry
• Check provider status page
• Try different model

Retry Logic

claude-code-Go automatically retries:

• Rate limits (with exponential backoff)
• Timeouts (up to 3 attempts)
• Server errors (5xx)

Manual retry:

> That failed, please try again