Search skmtc…
⌘K
Docs
Sign in
APIs
Generators
Stacks
Projects
Docs
History
Schema
Filter operations
agent_workers
5
ai
9
ai_evals
14
get
/w/{workspace}/ai_evals/datasets/list
list eval datasets
post
/w/{workspace}/ai_evals/datasets/create
create an eval dataset
get
/w/{workspace}/ai_evals/datasets/get/{path}
get an eval dataset
post
/w/{workspace}/ai_evals/datasets/update/{path}
update an eval dataset
post
/w/{workspace}/ai_evals/datasets/delete/{path}
delete an eval dataset and all its cases
get
/w/{workspace}/ai_evals/cases/list/{path}
list the cases of an eval dataset
get
/w/{workspace}/ai_evals/subject_state
what the agent under test is right now
get
/w/{workspace}/ai_evals/run_payload
the run one iteration of an eval run answered, as its scorers read it
get
/w/{workspace}/ai_evals/scorer_defaults
what a new judge agent and a new script scorer are created from
get
/w/{workspace}/ai_evals/scorers/recent
list the scorers already in use in this workspace, most recent first
post
/w/{workspace}/ai_evals/experiments/run
run every case of a dataset as one immutable experiment
post
/w/{workspace}/ai_evals/experiments/collect
record what a run produced, so it outlives the jobs that produced it
get
/w/{workspace}/ai_evals/experiments/list_all
list every experiment, across datasets
get
/w/{workspace}/ai_evals/experiments/results/{path}
read an experiment's results, one row per case
amqp_trigger
8
app
43
asset
10
audit
2
azure_trigger
13
capture
7
concurrencyGroups
4
config
10
data_metric
1
dbt
3
documentation
2
draft
7
email_trigger
8
favorite
2
flow
21
flow_conversations
4
folder
11
gcp_trigger
14
Git Sync
13
granular_acl
3
group
19
health
2
helpers
24
http_trigger
9
hubPublish
20
indexSearch
7
input
6
integration
2
job
104
kafka_trigger
10
mcp
1
mcp_oauth
3
metrics
3
mqtt_trigger
8
native_trigger
14
nats_trigger
8
npm_proxy
6
oauth
13
oidc
2
openapi
2
path_autocomplete
1
postgres_trigger
20
raw_app
1
resource
31
schedule
10
script
37
service_logs
2
setting
63
settings
4
sqs_trigger
8
teams
1
token
1
trash
5
trigger
3
user
65
variable
10
volume
4
websocket_trigger
8
worker
13
workspace
144
workspace_dependencies
5
workspace_integration
8
ai_evals
Copy Markdown
Copied
Copy failed
Changes to record what a run produced, so it outlives the jobs that produced it
post
/w/{workspace}/ai_evals/experiments/collect
Aug 24, 2026
5afc9f1a93f5
Revision changes
info
endpoint added