Agent Insight Monitors

createRun

POST /agent_insight_monitors/{monitor_id}/runs

Start an Agent Insights run for a monitor.

Header parameters

Foundry-Features enum required

A feature flag opt-in required when using preview operations or modifying persisted preview resources.

AgentInsights=V1Preview

Operation-Id string

Client-generated unique ID for idempotent retries. When absent, the server creates the job unconditionally.

Path parameters

monitor_id string required

The identifier of the monitor.

Query parameters

api-version string required

The API version to use for this operation.

Request Body

Run inputs. Send an empty object to use the default 168-hour lookback window.

Content-Type: application/json

lookback_hours number

Optional finite positive number of hours of trace history to analyze, up to 2,160. Defaults to 168.

Responses

The request has succeeded and a new resource has been created as a result.

Headers

Operation-Location string required

URL to poll for the job status.

Location string required

URL of the created job resource.

Body

Content-Type: application/json

id string required

Server-assigned unique identifier.

inputs AgentInsightRunCreate

Inputs used when creating an agent insight run.

lookback_hours number

Optional finite positive number of hours of trace history to analyze, up to 2,160. Defaults to 168.

result AgentInsightRunResult

Result statistics produced when an agent insight run succeeds.

traces_in_window integer required

The number of traces in the analyzed time window.

traces_analyzed integer required

The number of traces analyzed by the run.

insights_created integer required

The number of insights created by the run.

insights_updated integer required

The number of insights updated by the run.

insights_reopened integer required

The number of insights reopened by the run.

token_usage AgentInsightTokenUsage required

Token usage for an Agent Insights run.

input_tokens integer required

The number of input tokens used by the run.

output_tokens integer required

The number of output tokens used by the run.

cached_tokens integer

The number of input tokens served from cache.

total_tokens integer required

The total number of tokens used by the run.

status string | enum required

Extensible status values shared by Foundry jobs.

Any of the following:

string

enum

queuedin_progresssucceededfailedcancelled

error OpenAI.Error

code string | null required

message string required

param string | null

type string

misalignment OpenAI.MisalignmentErrorDetailsResource

review_target string | null

error_type string | enum

Any of the following:

string

enum

potentially_unintended_data_transferpotentially_unintended_data_accesspotentially_unintended_destructive_activityother

detailed_explanation string

The public explanation for this block.

steer OpenAI._MisalignmentSteer

message string required

The public continuation instruction.

details OpenAI.Error[]

additionalInfo object

debugInfo object

monitor_id string required

The Agent Insights monitor this run belongs to.

agent_name string required

The agent whose traces are analyzed by this run.

trigger string | enum required

The trigger that started an agent insight run.

Any of the following:

string

enum

on_demandscheduled

created_at integer required

updated_at integer required

window_start integer required

window_end integer required

started_at integer

completed_at integer

model_deployment_name string required

The model deployment used to analyze traces for this run.