Get Failures

GET

/admin/eval/failures

• Production
• Local development

Return the most recent failing eval runs, newest first.

Args: limit: Max rows (1..500). session: Async DB session. _admin: Operator auth gate.

Returns: List of :class:FailureRow sorted by run_at descending.

Authorizations

• ApiKeyAuth

Parameters

Query Parameters

limit

Max rows.

integer

default: 50 >= 1 <= 500

Max rows.

Responses

200

Successful Response

Array<object>

Single failing eval run surfaced in the Recent Failures table.

object

model_used

required

Any of:

string

string

null

null

output_snippet

required

Any of:

string

string

null

null

role

required

string

run_at

required

string format: date-time

score

required

Any of:

number

number

null

null

task_id

required

Any of:

string

string

null

null

422

Validation Error

object

detail

Array<object>

object

ctx

object

input

loc

required

Array

msg

required

string

type

required

string