Schema introspection
Every command below with arguments supports two read-only flags for tooling and AI agents:--show-file-schema: print the JSON Schema for the full request object (the format--filenameaccepts).--show-argument-schema <flag-name>: print the JSON Schema for one argument’s value. Pass the kebab flag name without the leading--(e.g.type, not--type).
list
List all metrics
Usage
Options
| Flag | Type | Required | Description |
|---|---|---|---|
--id | string | no | Filter by specific metric ID |
--type | string | no | Filter by metric type |
--show-file-schema, --show-argument-schema <flag-name>, and --filename. See Schema introspection for details.
create
Create a new metric
Usage
Options
| Flag | Type | Required | Description |
|---|---|---|---|
--criteria | string | yes | criteria |
--name | string | yes | name |
--type | string | yes | type Allowed: PYTHON, LLM, HUMAN, COMPOSITE. |
--categories | json | no | categories |
--child-metrics | json | no | child_metrics |
--description | string | no | description |
--enabled-in-prod / --no-enabled-in-prod | boolean | no | enabled_in_prod |
--filters | json | no | filters |
--model-name | string | no | model_name |
--model-provider | string | no | model_provider |
--needs-ground-truth / --no-needs-ground-truth | boolean | no | needs_ground_truth |
--return-type | string | no | return_type Allowed: float, boolean, string, categorical. |
--sampling-percentage | number | no | sampling_percentage |
--scale | number | no | scale |
--threshold | json | no | threshold |
--show-file-schema, --show-argument-schema <flag-name>, and --filename. See Schema introspection for details.
update
Update an existing metric
Update a metric’s editable fields. Only fields included in the request body are modified.
Usage
Options
| Flag | Type | Required | Description |
|---|---|---|---|
--metric-id | string | yes | The unique identifier of the metric to update |
--categories | json | no | categories |
--child-metrics | json | no | child_metrics |
--criteria | string | no | criteria |
--description | string | no | description |
--enabled-in-prod / --no-enabled-in-prod | boolean | no | enabled_in_prod |
--filters | json | no | filters |
--model-name | string | no | model_name |
--model-provider | string | no | model_provider |
--name | string | no | name |
--needs-ground-truth / --no-needs-ground-truth | boolean | no | needs_ground_truth |
--return-type | string | no | return_type Allowed: float, boolean, string, categorical. |
--sampling-percentage | number | no | sampling_percentage |
--scale | number | no | scale |
--threshold | json | no | threshold |
--type | string | no | type Allowed: PYTHON, LLM, HUMAN, COMPOSITE. |
--show-file-schema, --show-argument-schema <flag-name>, and --filename. See Schema introspection for details.
delete
Delete a metric
Usage
Options
| Flag | Type | Required | Description |
|---|---|---|---|
--metric-id | string | yes | The unique identifier of the metric to delete |
--show-file-schema, --show-argument-schema <flag-name>, and --filename. See Schema introspection for details.
run
Run a metric evaluation
Execute a metric on a specific event
Usage
Options
| Flag | Type | Required | Description |
|---|---|---|---|
--event | json | yes | event |
--metric | json | yes | metric |
--show-file-schema, --show-argument-schema <flag-name>, and --filename. See Schema introspection for details.