dbt Cloud CLI: dbt_cloud.yml not found / not authenticated

dbt Cloud CLI: dbt_cloud.yml not found / not authenticated

TL;DR: the CLI reads its credentials from dbtcloud.yml, which lives in your project dir and must contain a valid api key plus the right account, project, and environment ids. Regenerate it from the dbt Cloud UI (Develop > configure the IDE / download dbtcloud.yml), drop it in the project root, and rerun from that directory. If the file exists but auth still fails, the api key or ids are stale.

dbt_cloud.yml not found / not authenticated

Steps

1. In the dbt Cloud UI, go to your profile settings and generate a fresh api key. Then download or copy the generated dbt_cloud.yml for your environment.

1. Place dbt_cloud.yml in the root of the project you run the CLI from. The CLI looks in the current directory, not ~/.dbt.

1. Run dbt-cloud --version or a lightweight command from that directory. Expected: it authenticates instead of complaining about the config file.

1. Still failing: verify accountid, projectid, and environment_id in the file match the dbt Cloud URL you log into, and that the api key has not been revoked.

When this applies

• the dbt Cloud CLI (the dbt-cloud binary) reporting a missing or invalid dbt_cloud.yml
• not authenticated errors before any dbt command runs against dbt Cloud
• a fresh machine where the CLI was installed but never configured

When it doesn't

• dbt Core's profiles.yml — the Cloud CLI does not read it
• warehouse credential errors inside a run (the Cloud auth already succeeded)
• SSO browser flows for the dbt Cloud web UI itself

Compatibility

dbt Cloud CLI; dbt_cloud.yml format from the official dbt Cloud CLI configuration docs.

Variant phrasings

• dbt cloud cli dbt_cloud.yml not found
• dbt cloud cli not authenticated
• dbt-cloud configure authentication

Root cause

Unlike dbt Core, the Cloud CLI authenticates to the dbt Cloud API, not to your warehouse directly. dbt_cloud.yml is the whole credential bundle (host, ids, api key), so a missing file or a revoked key fails every command up front.

Edge cases

• running from a subdirectory means the CLI cannot find dbt_cloud.yml; cd to the project root first
• revoked api keys after a team member leaves produce the same not authenticated message
• the generated file embeds your personal api key; do not commit it to git