Environment Variables

Variable	Required	Description
`LOOMLOOM_SERVER`	Legacy/new-profile bootstrap	Full HTTPS base URL paired with the global Token
`LOOMLOOM_TOKEN`	Legacy/new-profile bootstrap	Global Token; valid only when bound to the same `LOOMLOOM_SERVER`
`LOOMLOOM_TOKEN_<PROFILE>`	Recommended after verification	Profile-specific Token variable generated and returned by `doctor`
`LOOMLOOM_VERBOSE`	No	Go-style boolean that enables debug logs on stderr

Do not guess a profile-specific variable. Register the exact Server and Token pair, then read token_env from the successful result:

loomloom doctor --server "<exact-server>" --token "<matching-token>" --output json

Precedence

For the Server, resolution order is:

--server
Active verified Server profile
LOOMLOOM_SERVER when no profile state exists

For the token, resolution order is:

--token
Active profile’s generated LOOMLOOM_TOKEN_<PROFILE> variable
Browser credential saved by loomloom login for that profile
LOOMLOOM_TOKEN only when it is bound to the selected Server

Legacy BATCHJOB_SERVER and BATCHJOB_TOKEN are not part of the v0.2.1 profile resolution path. New setups should use verified profiles and their generated Token variables.

Bootstrap a new profile

Pass the Server and Token explicitly to the first verification rather than relying on temporary global variables:

loomloom doctor \
  --server "https://loomloom.cogfoundry.ai/loom/v1" \
  --token "<cogfoundry-api-token>" \
  --output json

If the result returns next_action: persist_token, store the Token under the returned token_env. Do not commit shell profiles or environment files containing real tokens. For profile management, see Server profiles.

loomloom CLI

TemplateSpec

loomloom API

loomloom API endpoints

Precedence

Bootstrap a new profile

​Precedence

​Bootstrap a new profile

Precedence

Bootstrap a new profile