Command reference

Full command tree for the nora CLI. Each block links to a deeper page or is self-explanatory from the description.

`nora auth`

Sign-in and workspace management. See Auth.

Command	Description
`auth login`	OAuth browser sign-in
`auth logout`	Delete local credentials
`auth status`	Show active workspace and token
`auth workspaces list`	Refresh + print saved workspaces
`auth workspaces switch <id>`	Change active workspace
`auth workspaces current`	Print active workspace id

`nora flows`

Manage Flows. See Flows.

Command	Description
`flows list`	Every Flow in the workspace
`flows get <slug>`	Full Flow payload
`flows snapshot`	Just the editable payload
`flows apply --file <path> [<slug>]`	Upsert from JSON
`flows delete <slug>`	Remove a Flow
`flows use <slug>`	Pin active Flow
`flows current`	Print active Flow
`flows publish [--note "..."]`	Ship the Draft
`flows rollback <version-id>`	Move published pointer back
`flows settings get`	Read Flow settings
`flows settings update [flags]`	Patch settings
`flows versions list`	Version history
`flows versions get <id>`	Version metadata
`flows versions restore <id>`	Copy version into Draft

`nora agents`

Edit Agent blocks in the active Flow.

Command	Description
`agents create --name <n>`	New Agent block
`agents update <agent-id> [flags]`	Patch Agent fields
`agents edit <agent-id>`	Open in `$EDITOR`
`agents delete <agent-id>`	Remove Agent
`agents sources list <agent-id>`	Data sources on the Agent
`agents sources attach <agent-id> ...`	Attach a source
`agents sources detach <agent-id> ...`	Remove a source
`agents memories list <agent-id>`	Memory spaces on the Agent
`agents memories attach <agent-id> ...`	Attach a memory space
`agents memories detach <agent-id> ...`	Detach

Common agents update --set-* flags:

Identity: --set-name, --set-description, --set-prompt, --set-prompt-file
Model: --set-model, --set-mode
Limits: --set-max-turns, --set-max-tokens, --set-max-cost
Approvals: --set-approval on|off, --set-cost-approval USD|off

`nora tools`

Edit Tool blocks (HTTP or code).

Command	Description
`tools create --name <n>`	New Tool block
`tools update <tool-id> [flags]`	Patch Tool fields
`tools delete <tool-id>`	Remove
`tools params list/add/update/remove/reorder`	Manage parameters
`tools headers list/add/remove`	Manage HTTP headers
`tools verify-db <tool-id>`	Test a stored DB connection

Runtime kind: --set-runtime-kind http|inline. Inline supports --set-inline-lang javascript|python|sql.

`nora actions`

Same shape as tools, minus params (Actions fire after the Agent — no LLM-picked args).

Command	Description
`actions create --name <n>`	New Action block
`actions update <action-id>`	Patch
`actions delete <action-id>`	Remove
`actions headers list/add/remove`	Manage HTTP headers

`nora schema`

Command	Description
`schema show`	Print the Flow JSON schema

Handling secrets

Any flag that accepts a sensitive value takes:

? — interactive prompt (masked for secret fields)
- — read one line from stdin (pipe-friendly)
literal string — used as-is, with a shell-history warning for secret fields

Example:

nora tools headers add <tool-id> --name Authorization --value ?
# Prompts: "Authorization: " (masked)

echo "$MY_URL" | nora tools update <tool-id> --set-inline-database-url -

Global flags

Every subcommand accepts:

--base-url <url> — override server
--workspace <id> — override active workspace for this command
--flow <slug> — override active Flow for this command
-h, --help — print help

​nora auth

​nora flows

​nora agents

​nora tools

​nora actions

​nora schema

​Handling secrets

​Global flags