Actions
| Action | Description |
|---|---|
list | List all evals for a task |
create | Create a new eval |
get | Get an eval by ID |
update | Update an eval |
delete | Delete an eval |
Parameters
list
| Parameter | Type | Required | Description |
|---|---|---|---|
taskId | string | yes | Task ID to list evals for |
create
| Parameter | Type | Required | Description |
|---|---|---|---|
taskId | string | yes | Task ID to add the eval to |
label | string | yes | Human-readable label for the eval |
assert | string | yes | Assertion to evaluate (natural language or expression) |
weight | number | no | Relative weight for scoring (default 1) |
get
| Parameter | Type | Required | Description |
|---|---|---|---|
id | string | yes | Eval ID |
update
| Parameter | Type | Required | Description |
|---|---|---|---|
id | string | yes | Eval ID |
label | string | no | Updated label |
assert | string | no | Updated assertion |
weight | number | no | Updated weight |
delete
| Parameter | Type | Required | Description |
|---|---|---|---|
id | string | yes | Eval ID |