Skip to main content
When something goes wrong in a Finkkle Spaces conversation, plugin call, or automation, Spaces classifies the failure into one of the error types below. Each error is shown with a short code and a themed exoplanet name so you can recognize it at a glance in logs and toast messages. Use this page to identify what an error means and what to try next.

​
Quick reference

​
404 Dagon: Not found

The resource Spaces tried to reach does not exist, or is not visible to your account. Common causes
  • A conversation, artifact, or file was deleted or moved.
  • A plugin is pointing at a URL, repo, or endpoint that no longer exists.
  • A shared link was revoked.
  • A custom plugin is calling a path that was renamed on the upstream API.
How to fix
  • Refresh the page or reopen the conversation to reload the latest state.
  • Check that the referenced file, artifact, or repo still exists in your workspace.
  • For custom plugins, verify the endpoint path in your plugin config.
  • If you followed a shared link, ask the owner to reshare it.

​
403 / 401 Pegasi: Forbidden or unauthorized

Spaces reached the resource but was not allowed to access it. Common causes
  • Your session expired and you need to sign in again.
  • A connected integration (Slack, GitHub, Google, etc.) had its token revoked.
  • You are trying to access a Space, artifact, or plugin you have not been granted access to.
  • A custom plugin is missing an API key or is using an invalid one.
How to fix
  • Sign out and back into Finkkle.
  • Reconnect the affected integration from your plugin settings.
  • Ask the Space owner to invite you or grant the right role.
  • For custom plugins, rotate or re-enter the API key and confirm the required scopes.

​
408 Corot: Timeout

The request took too long and was aborted before it completed. Common causes
  • A long-running research or code task exceeded its execution window.
  • A plugin called a slow upstream service.
  • The client tab was backgrounded or the network stalled mid-request.
  • You manually cancelled the operation.
How to fix
  • Retry the request. Transient slowness usually clears on the next attempt.
  • Narrow the scope of the task (fewer files, shorter research prompt, smaller code diff).
  • For automations, split the task into smaller steps.
  • If a specific plugin times out repeatedly, check its status in the plugin catalog.

​
429 Lich: Rate limit

Too many requests were sent in a short window, either by you or by a plugin acting on your behalf. Common causes
  • Rapid retries after a failure.
  • An automation running on an aggressive schedule.
  • A custom plugin looping over a large dataset without backoff.
  • Hitting an upstream provider’s per-minute quota.
How to fix
  • Wait a minute and try again.
  • Reduce the frequency of the triggering automation.
  • Batch requests in custom plugins and add exponential backoff.
  • If you consistently hit the limit, review your plan’s usage tier.

​
400 Osiris: Bad request

Spaces sent a request that the target rejected as malformed. Common causes
  • A plugin was called with missing or invalid parameters.
  • A file upload exceeded size or format limits.
  • A malformed URL or search query.
  • A custom plugin schema does not match what the upstream API expects.
How to fix
  • Re-check the arguments you passed to the plugin or command.
  • For uploads, confirm the file type and size are supported.
  • For custom plugins, validate your request schema against the upstream API docs.

​
503 / 502 / 504 Centauri: Service unavailable

An upstream service Spaces depends on is down, overloaded, or unreachable through its gateway. Common causes
  • A model provider or integration is having an outage.
  • Temporary gateway or load balancer issues.
  • Maintenance on a connected service.
How to fix
  • Retry in a few minutes.
  • Check the status page for the affected provider or integration.
  • Switch to a different mode or model if the outage is provider-specific.

​
500 Kepler: Server error

Something failed inside Finkkle or a downstream service in a way that was not a client mistake. Common causes
  • An unhandled exception in a plugin or model call.
  • A bug in a recent release.
  • Corrupt state on a specific conversation or artifact.
How to fix
  • Retry the action. Many 500s are transient.
  • Start a fresh conversation to see if the error is scoped to one thread.
  • If it persists, report the conversation ID to support so it can be traced.

​
Network Wasp: Network or connection failure

The request never reached its destination because the browser or device could not complete the connection. Common causes
  • You lost Wi-Fi or cellular connectivity.
  • A VPN, firewall, or corporate proxy blocked the request.
  • DNS failure or captive portal.
  • Browser extensions blocking fetches to Finkkle domains.
How to fix
  • Confirm you have a working internet connection.
  • Disable VPN, proxy, or ad-blocking extensions and retry.
  • Allow finkkle.com and its subdomains through your firewall.
  • Try a different browser or network to isolate the issue.

​
UNKNOWN: Fallback

Spaces could not match the failure to any known category, so it shows a generic “Something went wrong” message. What to do
  • Retry the action once.
  • Reload the Space and try again in a new conversation.
  • If the error persists, contact support with the timestamp and, if available, the conversation ID. Unknown errors are the most useful to report because they often indicate a gap in Spaces’ error classification.

​
Reporting a persistent error

When an error keeps recurring, include the following when you contact support:
  • The error code and exoplanet name shown in the UI.
  • The conversation or automation ID.
  • The plugin involved, if any.
  • The approximate timestamp and your timezone.
  • What you were trying to do when it happened.