Sume CLI not working? Fixes for login, API base, jobs, and media

When the Sume CLI is not working, run four read-only checks first: sume version, sume auth status, sume doctor --agent --json, and sume account get --json. Then match the symptom: no API key configured, the wrong API base, a job that has not finished, a generation command that does not exist, or a media URL Sume cannot fetch.

The fixes come from the CLI Troubleshooting and Security pages, read on 2026-09-26. The CLI still works, but the docs say it is not part of the primary path today. Install and first login are in the CLI tutorial.

Which CLI problem do I have?

sume doctor --agent --json inspects local readiness without calling the API, so it is safe to run first. The table maps each symptom the docs cover to its fix.

How do I fix a missing key or the wrong API base?

For a missing key, run sume login and confirm with sume auth status. If the key still seems missing, or calls go to the wrong place, check where the CLI reads its settings:

• Auth can come from environment variables or from local config, ~/.sume-com/config.json by default. The docs' issue checklist asks which one a failing command used, so check both.
• SUME_CONFIG_DIR overrides the local config directory, so check it when the CLI does not see a login you already completed.
• SUME_API_BASE_URL defaults to https://api.sume.com/v1, the production base. If calls go somewhere else, check that variable and the doctor output.
• In CI, set a key from the API Keys dashboard with sume auth setup --api-key "$SUME_API_KEY" or the SUME_API_KEY variable; the CI guide covers the unattended setup.

Why does my job have no result yet?

Jobs are asynchronous. Poll the status, and fetch the result only after the job completes. If a local wait timed out, inspect the job before you submit another paid job: sume jobs watch polls until the job is terminal or times out, and recovery commands replace resubmitting paid generation.

sume jobs status <job_id> --agent --json
sume jobs result <job_id> --agent --json
sume jobs watch <job_id>

Why is there no sume video command?

The shipped CLI has no sume image, sume video, or sume music subcommand. Its docs send that work to the Developer API, pointing at Image 1.0, Video 1.0, and Music 1.0. Image 1.0 and Video 1.0 are retiring soon, so new integrations use POST /v1/images or POST /v1/videos with sume/auto, as the migration guide shows. Music 1.0 is retiring gradually in favor of POST /v1/music-router/generate. Either way, sume jobs status and sume jobs result still recover the job.

Why was my media URL rejected?

Avatar media fields expect public HTTPS image URLs. For Avatar 1.0 photo input, pass --type photo --image-url https://...; for Avatar Video, --product-image https://... or --scene-image-url https://.... The docs list four common causes:

• The URL is not HTTPS.
• The URL points to localhost or a private network.
• The response is not an image.
• The URL needs cookies, auth headers, or a short-lived signature that expires before Sume can fetch it.

What should I include when I report a CLI issue?

Include the command name and flags with secrets redacted, the sume version output, the sanitized error code and message, the request id if present, the job id if the issue is job-specific, and whether auth came from environment variables or local config.

Leave out API keys, signed URLs, private media URLs, raw provider payloads, emails, and workspace or user ids. When you share results, summarize media counts and file types, use local filenames instead of full remote URLs, and redact query strings.