Skip to content

Generation ​

api.md §3's generation record. Built from the row field by field, never from_attributes: the row also carries the image digest (architecture.md §2.15).

Example Usage ​

typescript
import { Generation } from "@meetkai/mka1/models/components";

let value: Generation = {
  id: "<id>",
  state: "Hawaii",
  reason: {},
  generator: {
    id: "<id>",
    version: "<value>",
  },
  params: {
    "key": "<value>",
    "key1": "<value>",
    "key2": "<value>",
  },
  paramsHash: "<value>",
  output: {
    repo: "<value>",
    revision: "<value>",
  },
  inputs: {
    "key": "<value>",
  },
  counts: {
    "key": 246883,
    "key1": 266259,
  },
  llm: {
    models: [],
  },
  limits: {
    maxRuntimeHours: 9692.81,
    maxCost: 93.78,
    maxLlmCalls: 244736,
  },
  usage: {
    "key": "<value>",
    "key1": "<value>",
    "key2": "<value>",
  },
  job: null,
  name: "<value>",
  derivedFrom: [],
  createdBy: {
    userId: "<id>",
    teamId: "<id>",
    apiKeyId: "<id>",
  },
  createdAt: new Date("2026-11-10T14:39:15.476Z"),
  startedAt: null,
  terminalAt: null,
};

Fields ​

FieldTypeRequiredDescription
idstring✔️N/A
statestring✔️N/A
reasonRecord<string, any>✔️N/A
generatorcomponents.GeneratorRef✔️N/A
paramsRecord<string, any>✔️N/A
paramsHashstring✔️N/A
outputcomponents.OutputRef✔️N/A
inputsRecord<string, string>✔️N/A
countsRecord<string, number>✔️N/A
llmcomponents.LlmRef✔️N/A
limitscomponents.Limits✔️A run's budget: the generator's default_limits, or the effective ones as ctx.limits.

No defaults. Every number here is a bound the platform enforces and a caller can be billed
against, so a generator that forgets to state one should not silently inherit a guess.
usageRecord<string, any>✔️N/A
jobcomponents.JobRef✔️N/A
namestring✔️N/A
derivedFromstring[]✔️N/A
createdBycomponents.CreatedBy✔️N/A
createdAtDate✔️N/A
startedAtDate✔️N/A
terminalAtDate✔️N/A