Explore the contract
Switch question types and inspect complete JSON requests and responses.
Request and response explorer
All three examples evaluate the same incident report. Select a primitive to compare its criteria and answer fields.
ILLUSTRATIVE EXAMPLE
Select an option from a set you define.
1{
2 "model": "example-model",
3 "state": "The nightly export stopped at 02:14. New reports are missing.",
4 "questions": {
5 "result": {
6 "type": "choice",
7 "instructions": "Which team should investigate?",
8 "criteria": {
9 "operations": "Failed jobs, missing data, or service outages",
10 "accounts": "Access, invitations, or account settings",
11 "other": "Requests outside these categories"
12 }
13 }
14 }
15}Example probabilities
operations
92%
accounts
5%
other
3%
Ask several questions
Place questions under different names to share the incident report. Each answer is evaluated against that state independently.
1{
2 "model": "example-model",
3 "state": "The nightly export stopped at 02:14. New reports are missing.",
4 "questions": {
5 "team": {
6 "type": "choice",
7 "instructions": "Which team should investigate?",
8 "criteria": {
9 "operations": "Failed jobs, missing data, or service outages",
10 "accounts": "Access, invitations, or account settings",
11 "other": "Requests outside these categories"
12 }
13 },
14 "impact": {
15 "type": "score",
16 "instructions": "How much does this affect the reporting workflow?",
17 "criteria": [
18 "No impact",
19 "Delayed work",
20 "Work blocked"
21 ]
22 },
23 "failed_job": {
24 "type": "noul",
25 "instructions": "Does the report describe a failed automated job?"
26 }
27 }
28}