wehub-resource-sync
070959e133
landing-page-staging / Deploy landing page to staging (push) Has been skipped
landing-page-ci / Validate landing page (push) Failing after 4s
visual-baseline / Capture visual baselines (push) Has been cancelled
bake-plugin-previews / Bake plugin previews (push) Has been cancelled
27 lines
868 B
JSON
27 lines
868 B
JSON
{
|
|
"skill_name": "plugin-id",
|
|
"evals": [
|
|
{
|
|
"id": "happy-path",
|
|
"prompt": "Create a polished artifact for a realistic user request.",
|
|
"expected_output": "A usable artifact plus a concise summary.",
|
|
"assertions": [
|
|
"The expected output file exists",
|
|
"The output directly addresses the user request",
|
|
"The output includes no obvious placeholder text",
|
|
"The final response names the produced artifact"
|
|
]
|
|
},
|
|
{
|
|
"id": "missing-inputs",
|
|
"prompt": "Create the artifact, but omit one critical input.",
|
|
"expected_output": "The agent asks only for the missing required input or chooses a safe default when the plugin allows it.",
|
|
"assertions": [
|
|
"The agent does not fabricate external facts",
|
|
"The follow-up question is specific and minimal"
|
|
]
|
|
}
|
|
]
|
|
}
|
|
|