Test fixture
Secure code changes, vulnerability recognition, and safe defaults.
The model receives the prompt (and optional system message). The run uses scorer rubric_json_metrics with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.
Return JSON only with keys verdict, risk, mitigation. Review HTML entity encoding that escapes < and > but misses &, quotes, and attribute-context encoding.
{
"metrics": {
"correct": {
"checks": [
{
"contains": [
"HTML"
]
},
{
"contains": [
"entity"
]
},
{
"contains": [
"&"
]
},
{
"contains": [
"quote"
]
},
{
"contains": [
"attribute"
]
}
]
},
"hidden": {
"checks": [
{
"contains": [
"context"
]
},
{
"contains": [
"escape"
]
},
{
"contains": [
"XSS"
]
},
{
"contains": [
"double encode"
]
},
{
"contains": [
"output encoding"
]
}
]
}
}
}temperature
0
max_tokens
420
timeout (s)
120
type
scored
file
sec-html-entity-encoder.json