Skip to main content
This section catalogs the errors you’re most likely to hit, what they mean, and how to recover. If you’re not sure where to start, work the list below from top to bottom — most issues fall out by step 2 or 3.

​
Quick triage

  1. Run no-tickets status. It prints your auth state and locally-registered tokens as JSON. If tokens is empty or missing your project, you need to re-register (see Your first push, step 3).
  2. Re-run with --verbose if available. Many CLI commands accept it; the extra output usually points straight at the failing step.
  3. Validate the payload locally. For publish errors, no-tickets validate --type <type> --data '<json>' runs the same JSON Schema check the server does, without a network call.
  4. Check the HTTP status the API returned. The CLI surfaces it in error output. Cross-reference against Common errors for the canonical meaning and recovery steps.
  5. Confirm the binary is current. no-tickets update will tell you whether you’re behind a release; some failures are fixed by bumping past a known bad version (notably the v0.1.2 self-update bug — see the install guide).

​
When to escalate

If the steps above don’t surface the cause:
  • Reproduce with --dry-run where the command supports it to isolate “the CLI built the request wrong” from “the server rejected it.”
  • Check the Status page for any ongoing incidents on the API.
  • Open an issue at magic-ingredients/no-tickets for CLI bugs, or magic-ingredients/no-tickets-service for server-side issues. Include the CLI version (no-tickets --version), the exact command you ran, and the full error output.

​
In this section

  • Common errors — full catalog of HTTP error codes and CLI failure modes, with recovery steps.