Skip to main content
Manage benchmarks that group related tasks into a runnable test suite for your agent.

​
Actions

ActionDescription
listList all benchmarks in the organization
createCreate a new benchmark
getGet a benchmark by ID
updateUpdate a benchmark
deleteDelete a benchmark

​
Parameters

​
list

No parameters.

​
create

ParameterTypeRequiredDescription
namestringyesName of the benchmark
descriptionstringnoDescription of the benchmark
timeoutnumbernoDefault timeout in seconds for tasks in this benchmark
concurrencynumbernoMax concurrent task runs

​
get

ParameterTypeRequiredDescription
idstringyesBenchmark ID

​
update

ParameterTypeRequiredDescription
idstringyesBenchmark ID
namestringnoUpdated name
descriptionstring | nullnoUpdated description, or null to clear
timeoutnumber | nullnoUpdated timeout, or null to clear
concurrencynumber | nullnoUpdated concurrency, or null to clear

​
delete

ParameterTypeRequiredDescription
idstringyesBenchmark ID

​
Examples

​
Create a prior auth workflow benchmark

{
  "action": "create",
  "name": "Prior Auth Workflow",
  "description": "End-to-end prior authorization submission and tracking",
  "timeout": 300,
  "concurrency": 5
}

​
List all benchmarks

{
  "action": "list"
}

​
Update benchmark concurrency

{
  "action": "update",
  "id": "bm_abc123",
  "concurrency": 10
}