@meetkai/mka1 0.54.22 → 0.54.24

package/src/react-query/llmEvalsGetSuiteLeaderboard.ts

@@ -0,0 +1,163 @@
+/*
+ * Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
+ */
+
+import {
+  InvalidateQueryFilters,
+  QueryClient,
+  useQuery,
+  UseQueryResult,
+  useSuspenseQuery,
+  UseSuspenseQueryResult,
+} from "@tanstack/react-query";
+import {
+  ConnectionError,
+  InvalidRequestError,
+  RequestAbortedError,
+  RequestTimeoutError,
+  UnexpectedClientError,
+} from "../models/errors/httpclienterrors.js";
+import { ResponseValidationError } from "../models/errors/responsevalidationerror.js";
+import { SDKError } from "../models/errors/sdkerror.js";
+import { SDKValidationError } from "../models/errors/sdkvalidationerror.js";
+import * as operations from "../models/operations/index.js";
+import { useSDKContext } from "./_context.js";
+import {
+  QueryHookOptions,
+  SuspenseQueryHookOptions,
+  TupleToPrefixes,
+} from "./_types.js";
+import {
+  buildLlmEvalsGetSuiteLeaderboardQuery,
+  LlmEvalsGetSuiteLeaderboardQueryData,
+  prefetchLlmEvalsGetSuiteLeaderboard,
+  queryKeyLlmEvalsGetSuiteLeaderboard,
+} from "./llmEvalsGetSuiteLeaderboard.core.js";
+export {
+  buildLlmEvalsGetSuiteLeaderboardQuery,
+  type LlmEvalsGetSuiteLeaderboardQueryData,
+  prefetchLlmEvalsGetSuiteLeaderboard,
+  queryKeyLlmEvalsGetSuiteLeaderboard,
+};
+
+export type LlmEvalsGetSuiteLeaderboardQueryError =
+  | SDKError
+  | ResponseValidationError
+  | ConnectionError
+  | RequestAbortedError
+  | RequestTimeoutError
+  | InvalidRequestError
+  | UnexpectedClientError
+  | SDKValidationError;
+
+/**
+ * Get the coding-agent leaderboard for an eval suite
+ *
+ * @remarks
+ * Ranks the suite's completed coding-agent runs, one row per (agent, model, effort). Only runs that recorded an `agent_name` participate — a QA or transcription run of the same suite has no agent to attribute a score to and is not a board row. `accuracy` is the mean of `metric` (default `reward`) over the row's scored trials, and `accuracy_ci` is the 95% Wald interval around it; both are derived per request rather than stored, so rerunning a failed trial moves them instead of leaving a stale copy behind. `agent_version` is reported but is not part of the row identity, so an agent upgrade does not split a row in two. Rows are ordered by accuracy descending, nulls last. Pass `suite_version` when the suite has more than one, otherwise the board mixes task sets.
+ */
+export function useLlmEvalsGetSuiteLeaderboard(
+  request: operations.GetEvalSuiteLeaderboardRequest,
+  options?: QueryHookOptions<
+    LlmEvalsGetSuiteLeaderboardQueryData,
+    LlmEvalsGetSuiteLeaderboardQueryError
+  >,
+): UseQueryResult<
+  LlmEvalsGetSuiteLeaderboardQueryData,
+  LlmEvalsGetSuiteLeaderboardQueryError
+> {
+  const client = useSDKContext();
+  return useQuery({
+    ...buildLlmEvalsGetSuiteLeaderboardQuery(
+      client,
+      request,
+      options,
+    ),
+    ...options,
+  });
+}
+
+/**
+ * Get the coding-agent leaderboard for an eval suite
+ *
+ * @remarks
+ * Ranks the suite's completed coding-agent runs, one row per (agent, model, effort). Only runs that recorded an `agent_name` participate — a QA or transcription run of the same suite has no agent to attribute a score to and is not a board row. `accuracy` is the mean of `metric` (default `reward`) over the row's scored trials, and `accuracy_ci` is the 95% Wald interval around it; both are derived per request rather than stored, so rerunning a failed trial moves them instead of leaving a stale copy behind. `agent_version` is reported but is not part of the row identity, so an agent upgrade does not split a row in two. Rows are ordered by accuracy descending, nulls last. Pass `suite_version` when the suite has more than one, otherwise the board mixes task sets.
+ */
+export function useLlmEvalsGetSuiteLeaderboardSuspense(
+  request: operations.GetEvalSuiteLeaderboardRequest,
+  options?: SuspenseQueryHookOptions<
+    LlmEvalsGetSuiteLeaderboardQueryData,
+    LlmEvalsGetSuiteLeaderboardQueryError
+  >,
+): UseSuspenseQueryResult<
+  LlmEvalsGetSuiteLeaderboardQueryData,
+  LlmEvalsGetSuiteLeaderboardQueryError
+> {
+  const client = useSDKContext();
+  return useSuspenseQuery({
+    ...buildLlmEvalsGetSuiteLeaderboardQuery(
+      client,
+      request,
+      options,
+    ),
+    ...options,
+  });
+}
+
+export function setLlmEvalsGetSuiteLeaderboardData(
+  client: QueryClient,
+  queryKeyBase: [
+    suiteId: string,
+    parameters: {
+      suiteVersion?: number | undefined;
+      metric?: string | undefined;
+      agent?: string | undefined;
+      model?: string | undefined;
+      limit?: number | undefined;
+      xOnBehalfOf?: string | undefined;
+    },
+  ],
+  data: LlmEvalsGetSuiteLeaderboardQueryData,
+): LlmEvalsGetSuiteLeaderboardQueryData | undefined {
+  const key = queryKeyLlmEvalsGetSuiteLeaderboard(...queryKeyBase);
+
+  return client.setQueryData<LlmEvalsGetSuiteLeaderboardQueryData>(key, data);
+}
+
+export function invalidateLlmEvalsGetSuiteLeaderboard(
+  client: QueryClient,
+  queryKeyBase: TupleToPrefixes<
+    [
+      suiteId: string,
+      parameters: {
+        suiteVersion?: number | undefined;
+        metric?: string | undefined;
+        agent?: string | undefined;
+        model?: string | undefined;
+        limit?: number | undefined;
+        xOnBehalfOf?: string | undefined;
+      },
+    ]
+  >,
+  filters?: Omit<InvalidateQueryFilters, "queryKey" | "predicate" | "exact">,
+): Promise<void> {
+  return client.invalidateQueries({
+    ...filters,
+    queryKey: [
+      "@meetkai/mka1",
+      "evals",
+      "getSuiteLeaderboard",
+      ...queryKeyBase,
+    ],
+  });
+}
+
+export function invalidateAllLlmEvalsGetSuiteLeaderboard(
+  client: QueryClient,
+  filters?: Omit<InvalidateQueryFilters, "queryKey" | "predicate" | "exact">,
+): Promise<void> {
+  return client.invalidateQueries({
+    ...filters,
+    queryKey: ["@meetkai/mka1", "evals", "getSuiteLeaderboard"],
+  });
+}

package/src/react-query/llmEvalsImportAgentRun.ts

@@ -0,0 +1,109 @@
+/*
+ * Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
+ */
+
+import {
+  MutationKey,
+  useMutation,
+  UseMutationResult,
+} from "@tanstack/react-query";
+import { SDKCore } from "../core.js";
+import { llmEvalsImportAgentRun } from "../funcs/llmEvalsImportAgentRun.js";
+import { combineSignals } from "../lib/primitives.js";
+import { RequestOptions } from "../lib/sdks.js";
+import {
+  ConnectionError,
+  InvalidRequestError,
+  RequestAbortedError,
+  RequestTimeoutError,
+  UnexpectedClientError,
+} from "../models/errors/httpclienterrors.js";
+import { ResponseValidationError } from "../models/errors/responsevalidationerror.js";
+import { SDKError } from "../models/errors/sdkerror.js";
+import { SDKValidationError } from "../models/errors/sdkvalidationerror.js";
+import * as operations from "../models/operations/index.js";
+import { unwrapAsync } from "../types/fp.js";
+import { useSDKContext } from "./_context.js";
+import { MutationHookOptions } from "./_types.js";
+
+export type LlmEvalsImportAgentRunMutationVariables = {
+  request: operations.ImportAgentEvalRunRequest;
+  options?: RequestOptions;
+};
+
+export type LlmEvalsImportAgentRunMutationData =
+  operations.ImportAgentEvalRunResponseBody;
+
+export type LlmEvalsImportAgentRunMutationError =
+  | SDKError
+  | ResponseValidationError
+  | ConnectionError
+  | RequestAbortedError
+  | RequestTimeoutError
+  | InvalidRequestError
+  | UnexpectedClientError
+  | SDKValidationError;
+
+/**
+ * Import a finished coding-agent run
+ *
+ * @remarks
+ * Uploads one completed harness job — the run and its trials — directly, so results never round-trip through external storage the way the Hugging Face importer does. Every `task_id` must already exist in the target suite version's manifest; unknown ids are rejected rather than invented, because a typo would otherwise import as a task the suite does not have and read as a model that never covered the version. The run is recorded as `completed` and is not executed: this is a write, not a run request, so no workflow is started. Trials that cover every task of the version get `task_ids: null`, which is what lets the run rank as a complete sweep. Trajectories are NOT part of this payload — a full 91-task run measured 8.3 MB of them — so upload each one against `PUT /evals/runs/{run_id}/samples/{sample_index}/trajectory` afterwards. Pass `idempotency_key` and a retried upload returns the first run instead of doubling the trials behind a board row.
+ */
+export function useLlmEvalsImportAgentRunMutation(
+  options?: MutationHookOptions<
+    LlmEvalsImportAgentRunMutationData,
+    LlmEvalsImportAgentRunMutationError,
+    LlmEvalsImportAgentRunMutationVariables
+  >,
+): UseMutationResult<
+  LlmEvalsImportAgentRunMutationData,
+  LlmEvalsImportAgentRunMutationError,
+  LlmEvalsImportAgentRunMutationVariables
+> {
+  const client = useSDKContext();
+  return useMutation({
+    ...buildLlmEvalsImportAgentRunMutation(client, options),
+    ...options,
+  });
+}
+
+export function mutationKeyLlmEvalsImportAgentRun(): MutationKey {
+  return ["@meetkai/mka1", "evals", "importAgentRun"];
+}
+
+export function buildLlmEvalsImportAgentRunMutation(
+  client$: SDKCore,
+  hookOptions?: RequestOptions,
+): {
+  mutationKey: MutationKey;
+  mutationFn: (
+    variables: LlmEvalsImportAgentRunMutationVariables,
+  ) => Promise<LlmEvalsImportAgentRunMutationData>;
+} {
+  return {
+    mutationKey: mutationKeyLlmEvalsImportAgentRun(),
+    mutationFn: function llmEvalsImportAgentRunMutationFn({
+      request,
+      options,
+    }): Promise<LlmEvalsImportAgentRunMutationData> {
+      const mergedOptions = {
+        ...hookOptions,
+        ...options,
+        fetchOptions: {
+          ...hookOptions?.fetchOptions,
+          ...options?.fetchOptions,
+          signal: combineSignals(
+            hookOptions?.fetchOptions?.signal,
+            options?.fetchOptions?.signal,
+          ),
+        },
+      };
+      return unwrapAsync(llmEvalsImportAgentRun(
+        client$,
+        request,
+        mergedOptions,
+      ));
+    },
+  };
+}

package/src/react-query/llmEvalsUploadSampleTrajectory.ts

@@ -0,0 +1,110 @@
+/*
+ * Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
+ */
+
+import {
+  MutationKey,
+  useMutation,
+  UseMutationResult,
+} from "@tanstack/react-query";
+import { SDKCore } from "../core.js";
+import { llmEvalsUploadSampleTrajectory } from "../funcs/llmEvalsUploadSampleTrajectory.js";
+import { combineSignals } from "../lib/primitives.js";
+import { RequestOptions } from "../lib/sdks.js";
+import * as components from "../models/components/index.js";
+import {
+  ConnectionError,
+  InvalidRequestError,
+  RequestAbortedError,
+  RequestTimeoutError,
+  UnexpectedClientError,
+} from "../models/errors/httpclienterrors.js";
+import { ResponseValidationError } from "../models/errors/responsevalidationerror.js";
+import { SDKError } from "../models/errors/sdkerror.js";
+import { SDKValidationError } from "../models/errors/sdkvalidationerror.js";
+import * as operations from "../models/operations/index.js";
+import { unwrapAsync } from "../types/fp.js";
+import { useSDKContext } from "./_context.js";
+import { MutationHookOptions } from "./_types.js";
+
+export type LlmEvalsUploadSampleTrajectoryMutationVariables = {
+  request: operations.UploadEvalSampleTrajectoryRequest;
+  options?: RequestOptions;
+};
+
+export type LlmEvalsUploadSampleTrajectoryMutationData =
+  components.EvalSampleTrajectoryUploadedObject;
+
+export type LlmEvalsUploadSampleTrajectoryMutationError =
+  | SDKError
+  | ResponseValidationError
+  | ConnectionError
+  | RequestAbortedError
+  | RequestTimeoutError
+  | InvalidRequestError
+  | UnexpectedClientError
+  | SDKValidationError;
+
+/**
+ * Upload an eval sample trajectory
+ *
+ * @remarks
+ * Stores one coding-agent trial's session, in whatever interchange format the harness produced (record it in `schema_version`, e.g. `ATIF-v1.7`). `steps` is stored verbatim and never validated, so a harness format change needs no gateway release. Separate from the run import because a trajectory averages ~92 KB and a whole run's worth would not fit in one request. Upserts, so re-sending after a timeout replaces the trajectory rather than failing. Sample indices are unique per (task, model), not per run, so pass `task_id` and/or `model` when the run covers more than one — an ambiguous index returns 400 rather than writing to the wrong trial. Rejects a payload whose `steps` and `logs` together exceed 16 MiB with 413.
+ */
+export function useLlmEvalsUploadSampleTrajectoryMutation(
+  options?: MutationHookOptions<
+    LlmEvalsUploadSampleTrajectoryMutationData,
+    LlmEvalsUploadSampleTrajectoryMutationError,
+    LlmEvalsUploadSampleTrajectoryMutationVariables
+  >,
+): UseMutationResult<
+  LlmEvalsUploadSampleTrajectoryMutationData,
+  LlmEvalsUploadSampleTrajectoryMutationError,
+  LlmEvalsUploadSampleTrajectoryMutationVariables
+> {
+  const client = useSDKContext();
+  return useMutation({
+    ...buildLlmEvalsUploadSampleTrajectoryMutation(client, options),
+    ...options,
+  });
+}
+
+export function mutationKeyLlmEvalsUploadSampleTrajectory(): MutationKey {
+  return ["@meetkai/mka1", "evals", "uploadSampleTrajectory"];
+}
+
+export function buildLlmEvalsUploadSampleTrajectoryMutation(
+  client$: SDKCore,
+  hookOptions?: RequestOptions,
+): {
+  mutationKey: MutationKey;
+  mutationFn: (
+    variables: LlmEvalsUploadSampleTrajectoryMutationVariables,
+  ) => Promise<LlmEvalsUploadSampleTrajectoryMutationData>;
+} {
+  return {
+    mutationKey: mutationKeyLlmEvalsUploadSampleTrajectory(),
+    mutationFn: function llmEvalsUploadSampleTrajectoryMutationFn({
+      request,
+      options,
+    }): Promise<LlmEvalsUploadSampleTrajectoryMutationData> {
+      const mergedOptions = {
+        ...hookOptions,
+        ...options,
+        fetchOptions: {
+          ...hookOptions?.fetchOptions,
+          ...options?.fetchOptions,
+          signal: combineSignals(
+            hookOptions?.fetchOptions?.signal,
+            options?.fetchOptions?.signal,
+          ),
+        },
+      };
+      return unwrapAsync(llmEvalsUploadSampleTrajectory(
+        client$,
+        request,
+        mergedOptions,
+      ));
+    },
+  };
+}

package/src/react-query/llmResponsesCreate.ts

@@ -0,0 +1,113 @@
+/*
+ * Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
+ */
+
+import {
+  MutationKey,
+  useMutation,
+  UseMutationResult,
+} from "@tanstack/react-query";
+import { SDKCore } from "../core.js";
+import {
+  CreateAcceptEnum,
+  llmResponsesCreate,
+} from "../funcs/llmResponsesCreate.js";
+import { combineSignals } from "../lib/primitives.js";
+import { RequestOptions } from "../lib/sdks.js";
+import {
+  ConnectionError,
+  InvalidRequestError,
+  RequestAbortedError,
+  RequestTimeoutError,
+  UnexpectedClientError,
+} from "../models/errors/httpclienterrors.js";
+import { ResponseValidationError } from "../models/errors/responsevalidationerror.js";
+import { SDKError } from "../models/errors/sdkerror.js";
+import { SDKValidationError } from "../models/errors/sdkvalidationerror.js";
+import * as operations from "../models/operations/index.js";
+import { unwrapAsync } from "../types/fp.js";
+import { useSDKContext } from "./_context.js";
+import { MutationHookOptions } from "./_types.js";
+
+export { CreateAcceptEnum } from "../funcs/llmResponsesCreate.js";
+
+export type LlmResponsesCreateMutationVariables = {
+  request: operations.CreateResponseRequest;
+  options?: RequestOptions & { acceptHeaderOverride?: CreateAcceptEnum };
+};
+
+export type LlmResponsesCreateMutationData = operations.CreateResponseResponse;
+
+export type LlmResponsesCreateMutationError =
+  | SDKError
+  | ResponseValidationError
+  | ConnectionError
+  | RequestAbortedError
+  | RequestTimeoutError
+  | InvalidRequestError
+  | UnexpectedClientError
+  | SDKValidationError;
+
+/**
+ * Create an agent-powered response with tool support
+ *
+ * @remarks
+ * Creates a new AI agent response using advanced language models with autonomous tool usage capabilities.
+ */
+export function useLlmResponsesCreateMutation(
+  options?: MutationHookOptions<
+    LlmResponsesCreateMutationData,
+    LlmResponsesCreateMutationError,
+    LlmResponsesCreateMutationVariables
+  >,
+): UseMutationResult<
+  LlmResponsesCreateMutationData,
+  LlmResponsesCreateMutationError,
+  LlmResponsesCreateMutationVariables
+> {
+  const client = useSDKContext();
+  return useMutation({
+    ...buildLlmResponsesCreateMutation(client, options),
+    ...options,
+  });
+}
+
+export function mutationKeyLlmResponsesCreate(): MutationKey {
+  return ["@meetkai/mka1", "responses", "create"];
+}
+
+export function buildLlmResponsesCreateMutation(
+  client$: SDKCore,
+  hookOptions?: RequestOptions & { acceptHeaderOverride?: CreateAcceptEnum },
+): {
+  mutationKey: MutationKey;
+  mutationFn: (
+    variables: LlmResponsesCreateMutationVariables,
+  ) => Promise<LlmResponsesCreateMutationData>;
+} {
+  return {
+    mutationKey: mutationKeyLlmResponsesCreate(),
+    mutationFn: function llmResponsesCreateMutationFn({
+      request,
+      options,
+    }): Promise<LlmResponsesCreateMutationData> {
+      const mergedOptions = {
+        ...hookOptions,
+        ...options,
+        fetchOptions: {
+          ...hookOptions?.fetchOptions,
+          ...options?.fetchOptions,
+          signal: combineSignals(
+            hookOptions?.fetchOptions?.signal,
+            options?.fetchOptions?.signal,
+          ),
+        },
+      };
+      return unwrapAsync(llmResponsesCreate(
+        client$,
+        request,
+        mergedOptions,
+      ));
+    },
+  };
+}

package/src/sdk/evals.ts

@@ -2,6 +2,7 @@
  * Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
  */
 
+import { llmEvalsAppendAgentTrials } from "../funcs/llmEvalsAppendAgentTrials.js";
 import { llmEvalsCancelRun } from "../funcs/llmEvalsCancelRun.js";
 import { llmEvalsCreateRun } from "../funcs/llmEvalsCreateRun.js";
 import { llmEvalsCreateSchedule } from "../funcs/llmEvalsCreateSchedule.js";
@@ -12,10 +13,14 @@ import { llmEvalsDeleteSchedule } from "../funcs/llmEvalsDeleteSchedule.js";
 import { llmEvalsDeleteSuite } from "../funcs/llmEvalsDeleteSuite.js";
 import { llmEvalsGetArtifacts } from "../funcs/llmEvalsGetArtifacts.js";
 import { llmEvalsGetRun } from "../funcs/llmEvalsGetRun.js";
+import { llmEvalsGetRunOverview } from "../funcs/llmEvalsGetRunOverview.js";
 import { llmEvalsGetSampleAudio } from "../funcs/llmEvalsGetSampleAudio.js";
+import { llmEvalsGetSampleTrajectory } from "../funcs/llmEvalsGetSampleTrajectory.js";
 import { llmEvalsGetSchedule } from "../funcs/llmEvalsGetSchedule.js";
 import { llmEvalsGetSuite } from "../funcs/llmEvalsGetSuite.js";
+import { llmEvalsGetSuiteLeaderboard } from "../funcs/llmEvalsGetSuiteLeaderboard.js";
 import { llmEvalsGetSuiteVersion } from "../funcs/llmEvalsGetSuiteVersion.js";
+import { llmEvalsImportAgentRun } from "../funcs/llmEvalsImportAgentRun.js";
 import { llmEvalsImportHistoricalResults } from "../funcs/llmEvalsImportHistoricalResults.js";
 import { llmEvalsListRuns } from "../funcs/llmEvalsListRuns.js";
 import { llmEvalsListSamples } from "../funcs/llmEvalsListSamples.js";
@@ -28,6 +33,7 @@ import { llmEvalsRerunFailedSamples } from "../funcs/llmEvalsRerunFailedSamples.
 import { llmEvalsRetryFailedRun } from "../funcs/llmEvalsRetryFailedRun.js";
 import { llmEvalsTriggerSchedule } from "../funcs/llmEvalsTriggerSchedule.js";
 import { llmEvalsUpdateSchedule } from "../funcs/llmEvalsUpdateSchedule.js";
+import { llmEvalsUploadSampleTrajectory } from "../funcs/llmEvalsUploadSampleTrajectory.js";
 import { ClientSDK, RequestOptions } from "../lib/sdks.js";
 import * as components from "../models/components/index.js";
 import * as operations from "../models/operations/index.js";
@@ -170,6 +176,23 @@ export class Evals extends ClientSDK {
     ));
   }
 
+  /**
+   * Get the coding-agent leaderboard for an eval suite
+   *
+   * @remarks
+   * Ranks the suite's completed coding-agent runs, one row per (agent, model, effort). Only runs that recorded an `agent_name` participate — a QA or transcription run of the same suite has no agent to attribute a score to and is not a board row. `accuracy` is the mean of `metric` (default `reward`) over the row's scored trials, and `accuracy_ci` is the 95% Wald interval around it; both are derived per request rather than stored, so rerunning a failed trial moves them instead of leaving a stale copy behind. `agent_version` is reported but is not part of the row identity, so an agent upgrade does not split a row in two. Rows are ordered by accuracy descending, nulls last. Pass `suite_version` when the suite has more than one, otherwise the board mixes task sets.
+   */
+  async getSuiteLeaderboard(
+    request: operations.GetEvalSuiteLeaderboardRequest,
+    options?: RequestOptions,
+  ): Promise<components.EvalSuiteLeaderboard> {
+    return unwrapAsync(llmEvalsGetSuiteLeaderboard(
+      this,
+      request,
+      options,
+    ));
+  }
+
   /**
    * Create an eval run
    *
@@ -289,6 +312,23 @@ export class Evals extends ClientSDK {
     ));
   }
 
+  /**
+   * Get an eval run overview
+   *
+   * @remarks
+   * Aggregate header for one run's trials view: trial counts, average reward, cost, and the input/output/cached token split. Its own route rather than fields on the run object because it scans the run's samples — folding it into `GET /evals/runs/{run_id}` would charge that scan to every run list and to the status polling that runs for the whole length of an eval. `trials.errored` overlaps `trials.failed` instead of partitioning it: a coding-agent trial can carry an agent timeout in `error` and still be completed and scored. `trials.retried` counts trials rerun through rerun-failed, read from the pre-rerun snapshot, since a rerun overwrites the sample in place.
+   */
+  async getRunOverview(
+    request: operations.GetEvalRunOverviewRequest,
+    options?: RequestOptions,
+  ): Promise<components.EvalRunOverview> {
+    return unwrapAsync(llmEvalsGetRunOverview(
+      this,
+      request,
+      options,
+    ));
+  }
+
   /**
    * List eval samples
    *
@@ -323,6 +363,40 @@ export class Evals extends ClientSDK {
     ));
   }
 
+  /**
+   * Get an eval sample trajectory
+   *
+   * @remarks
+   * Returns one coding-agent trial's session — the steps the agent took, verbatim in the harness's interchange format (see `schema_version`, e.g. `ATIF-v1.7`). Deliberately not part of the samples list: a trajectory averages ~92 KB and that list pages up to 500 rows. Sample indices are unique per (task, model), not per run, so pass `task_id` and/or `model` when the run covers more than one; an index that matches several trials returns 400 rather than guessing, and never falls back to another trial's session. Returns 404 when the sample does not exist or recorded no trajectory.
+   */
+  async getSampleTrajectory(
+    request: operations.GetEvalSampleTrajectoryRequest,
+    options?: RequestOptions,
+  ): Promise<operations.GetEvalSampleTrajectoryResponseBody> {
+    return unwrapAsync(llmEvalsGetSampleTrajectory(
+      this,
+      request,
+      options,
+    ));
+  }
+
+  /**
+   * Upload an eval sample trajectory
+   *
+   * @remarks
+   * Stores one coding-agent trial's session, in whatever interchange format the harness produced (record it in `schema_version`, e.g. `ATIF-v1.7`). `steps` is stored verbatim and never validated, so a harness format change needs no gateway release. Separate from the run import because a trajectory averages ~92 KB and a whole run's worth would not fit in one request. Upserts, so re-sending after a timeout replaces the trajectory rather than failing. Sample indices are unique per (task, model), not per run, so pass `task_id` and/or `model` when the run covers more than one — an ambiguous index returns 400 rather than writing to the wrong trial. Rejects a payload whose `steps` and `logs` together exceed 16 MiB with 413.
+   */
+  async uploadSampleTrajectory(
+    request: operations.UploadEvalSampleTrajectoryRequest,
+    options?: RequestOptions,
+  ): Promise<components.EvalSampleTrajectoryUploadedObject> {
+    return unwrapAsync(llmEvalsUploadSampleTrajectory(
+      this,
+      request,
+      options,
+    ));
+  }
+
   /**
    * Get eval run artifacts
    *
@@ -340,6 +414,40 @@ export class Evals extends ClientSDK {
     ));
   }
 
+  /**
+   * Import a finished coding-agent run
+   *
+   * @remarks
+   * Uploads one completed harness job — the run and its trials — directly, so results never round-trip through external storage the way the Hugging Face importer does. Every `task_id` must already exist in the target suite version's manifest; unknown ids are rejected rather than invented, because a typo would otherwise import as a task the suite does not have and read as a model that never covered the version. The run is recorded as `completed` and is not executed: this is a write, not a run request, so no workflow is started. Trials that cover every task of the version get `task_ids: null`, which is what lets the run rank as a complete sweep. Trajectories are NOT part of this payload — a full 91-task run measured 8.3 MB of them — so upload each one against `PUT /evals/runs/{run_id}/samples/{sample_index}/trajectory` afterwards. Pass `idempotency_key` and a retried upload returns the first run instead of doubling the trials behind a board row.
+   */
+  async importAgentRun(
+    request: operations.ImportAgentEvalRunRequest,
+    options?: RequestOptions,
+  ): Promise<operations.ImportAgentEvalRunResponseBody> {
+    return unwrapAsync(llmEvalsImportAgentRun(
+      this,
+      request,
+      options,
+    ));
+  }
+
+  /**
+   * Append trials to an open coding-agent run
+   *
+   * @remarks
+   * Adds finished trials to a run opened with `finalize: false`, so a job started from MKA1 can be watched while it runs instead of appearing complete out of nowhere at the end. Each trial pins its own `sample_index` and upserts on (task, model, sample_index): a harness cannot tell a timeout from a lost response, so re-sending a batch is safe and replaces rather than duplicates. The whole batch and its aggregate recompute land in one transaction, serialized per run. Counts, metrics and cost are recomputed from the stored rows on every call, never accumulated from the request, so a replayed batch cannot double them. Send `finalize: true` (usually with an empty `trials`) to close the run — until then it stays `in_progress` and off the leaderboards, which only count completed runs. Appending to an already-completed or cancelled run is rejected: that record is finished, and rewriting it would silently move a number someone may have already read.
+   */
+  async appendAgentTrials(
+    request: operations.AppendAgentEvalTrialsRequest,
+    options?: RequestOptions,
+  ): Promise<operations.AppendAgentEvalTrialsResponseBody> {
+    return unwrapAsync(llmEvalsAppendAgentTrials(
+      this,
+      request,
+      options,
+    ));
+  }
+
   /**
    * Import historical eval results from Hugging Face
    *