Interface: EvalTestResult
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:99
Result of a single test case evaluation.
Properties
actualOutput
actualOutput:
string
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:111
Actual agent output
costUsd?
optionalcostUsd:number
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:123
Estimated cost
error?
optionalerror:string
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:125
Error if any
expectedOutput?
optionalexpectedOutput:string
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:113
Expected output
latencyMs
latencyMs:
number
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:115
Latency in ms
metrics
metrics:
MetricValue[]
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:109
Individual metric scores
passed
passed:
boolean
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:105
Whether the test passed
score
score:
number
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:107
Overall score (0-1)
testCaseId
testCaseId:
string
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:101
Test case ID
testCaseName
testCaseName:
string
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:103
Test case name
timestamp
timestamp:
string
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:127
Timestamp
tokenUsage?
optionaltokenUsage:object
Defined in: packages/agentos/src/core/evaluation/IEvaluator.ts:117
Token usage
completionTokens
completionTokens:
number
promptTokens
promptTokens:
number
totalTokens
totalTokens:
number