Recipe catalog / log-line-kind
Classify what a log line reports
What does line report: an error, a warning, a lifecycle event, a handled request, a metric, or debug output?
You are grouping or filtering unstructured log lines from mixed sources and cannot rely on a level field being present or accurate.
Explore this recipe interactively ยท Source and implementation guide
Use log-line-kind in TypeScript
Install with npm install jev-recipes. Requires Node.js 22.9 or newer and ES modules. Set TYPESAFE_API_KEY in your server environment for live calls, which send input to TypeSafe and use API quota. See the installation guide.
import { logLineKind } from 'jev-recipes/log-line-kind';
const result = await logLineKind({
"line": "2026-09-23T14:02:11Z WARN db.pool: connection pool at 87% capacity, above the 80% threshold; new requests may queue",
"minConfidence": 0.8
});
console.log(result);
Input contract
| Field | Type | Needed |
|---|---|---|
| line | string | Required |
| minConfidence | number | Optional |
Full input and result schemas
{
"input": {
"$schema": "https://json-schema.org/draft/2020-12/schema",
"type": "object",
"properties": {
"line": {
"type": "string"
},
"minConfidence": {
"type": "number",
"minimum": 0,
"maximum": 1
}
},
"required": [
"line"
]
},
"result": {
"$schema": "https://json-schema.org/draft/2020-12/schema",
"type": "object",
"properties": {
"model": {
"type": "string"
},
"usage": {
"type": "object",
"properties": {
"input_tokens": {
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
},
"output_tokens": {
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
}
},
"required": [
"input_tokens",
"output_tokens"
],
"additionalProperties": false
},
"status": {
"type": "string",
"enum": [
"ready",
"review"
]
},
"verdict": {
"type": "string",
"enum": [
"error",
"warning",
"lifecycle",
"request",
"metric",
"debug",
"unclear"
]
},
"confidence": {
"type": "number",
"minimum": 0,
"maximum": 1
},
"probabilities": {
"type": "object",
"propertyNames": {
"type": "string",
"enum": [
"error",
"warning",
"lifecycle",
"request",
"metric",
"debug",
"unclear"
]
},
"additionalProperties": {
"type": "number",
"minimum": 0,
"maximum": 1
},
"required": [
"error",
"warning",
"lifecycle",
"request",
"metric",
"debug",
"unclear"
]
}
},
"required": [
"model",
"usage",
"status",
"verdict",
"confidence",
"probabilities"
],
"additionalProperties": false
}
}Saved example result
This hand-authored response demonstrates the contract. It is not a model accuracy measurement. Run it without an API key: npx jev-recipes demo log-line-kind.
{
"model": "demo-fixture",
"usage": {
"input_tokens": 0,
"output_tokens": 0
},
"status": "ready",
"verdict": "warning",
"confidence": 0.82,
"probabilities": {
"error": 0.03,
"warning": 0.82,
"lifecycle": 0.01,
"request": 0.01,
"metric": 0.1,
"debug": 0.02,
"unclear": 0.01
}
}
Evaluation evidence
No verified live accuracy measurement is available. Evaluate representative cases before using this decision in your workflow.
Use the evaluation guide to measure this decision on your own labeled cases.
Limitations
- Classifies one line from its wording. Lines that only make sense with neighbouring lines or a stack trace may come back unclear.
- Does not trust or ignore a printed level token by rule; it weighs the token against what the line describes.
- Timestamps, request IDs, and numeric values are not parsed. Extract them in application code.
Related recipes
- failure-kind: Use failure-kind once a line is known to report an error and you need to say what kind of failure it was.
- result-outcome: Use result-outcome to classify how a reported task or job ended, rather than what kind of log line it is.