hanzoai 2.2.10 → 2.2.11

package/dist/api/benchmark-api.d.ts

@@ -15,11 +15,15 @@ import globalAxios from 'axios';
 import { type RequestArgs, BaseAPI } from '../base';
 import type { Admission } from '../models';
 import type { BenchmarkCatalog } from '../models';
+import type { ClaimsOut } from '../models';
+import type { HistoryOut } from '../models';
 import type { Leaderboard } from '../models';
 import type { Pairing } from '../models';
 import type { Preset } from '../models';
 import type { PresetAccepted } from '../models';
 import type { PresetList } from '../models';
+import type { PutClaimsIn } from '../models';
+import type { PutClaimsOut } from '../models';
 import type { Suite } from '../models';
 /**
  * BenchmarkApi - axios parameter creator
@@ -33,6 +37,18 @@ export declare const BenchmarkApiAxiosParamCreator: (configuration?: Configurati
      * @throws {RequiredError}
      */
     getBenchmarkCatalog: (options?: RawAxiosRequestConfig) => Promise<RequestArgs>;
+    /**
+     * Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered. It answers the operator\'s question — what does this arena currently believe someone else reported, and did we ship that or fix it.  Effective values only. The history of a key lives in the append-only file and is not what this op is for; a list that returned every superseded row would make the common question the hard one.
+     * @summary Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered.
+     * @param {string} [benchmark] Benchmark filters to one benchmark id. Empty returns every benchmark.
+     * @param {string} [model] Model filters to one model. Empty returns every model.
+     * @param {string} [provider] Provider filters to one lab or leaderboard — the way to read what a single source claims across every model it covers.
+     * @param {string} [source] Source filters to one citation, which is the finest grain there is: a source is what makes two claims about one model independent rather than a restatement of each other.
+     * @param {string} [protocol] Protocol filters by HOW a claim was scored, so provider cards can be read apart from third parties running their own harness.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    getBenchmarkClaims: (benchmark?: string, model?: string, provider?: string, source?: string, protocol?: string, options?: RawAxiosRequestConfig) => Promise<RequestArgs>;
     /**
      * Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.  Pairing is what prevents the subset artifact — comparing one model\'s easy subset against another\'s full run — so n_common, not either arm\'s own coverage, is the number to read this by.  Both a and b are required. The benchmark defaults to gpqa_diamond.
      * @summary Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.
@@ -43,6 +59,15 @@ export declare const BenchmarkApiAxiosParamCreator: (configuration?: Configurati
      * @throws {RequiredError}
      */
     getBenchmarkCompare: (a: string, b: string, benchmark?: string, options?: RawAxiosRequestConfig) => Promise<RequestArgs>;
+    /**
+     * Returns each model\'s measured score per run over time, oldest first, with the change between runs.  This is the counterweight to a leaderboard: the board shows the latest run because that is what \"how good is it\" means, and a single latest number cannot distinguish a model that has always been strong from one that just improved, or from one that regressed after a provider changed something. Both matter for routing, and only one of them is visible on a board.  Runs with no id — attempts recorded before runs existed — group under the empty run, which is honestly what they are: one undated measurement.
+     * @summary Returns each model\'s measured score per run over time, oldest first, with the change between runs.
+     * @param {string} [benchmark] Benchmark is the catalog id to read, defaulting to gpqa_diamond.
+     * @param {string} [model] Model filters to one model. Empty returns every model measured.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    getBenchmarkHistory: (benchmark?: string, model?: string, options?: RawAxiosRequestConfig) => Promise<RequestArgs>;
     /**
      * Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.  The gap is the point of the arena; provider-reported claims have run materially hot against one standardized harness.  The two planes are NEVER blended, and that is the rule to read the rows by: a model we have measured but no vendor has claimed for shows published null, a model with only a claim shows measured null, and gap exists only where both do.  n is coverage and is not decoration: two measured numbers taken over different item counts are not comparable, so read the row\'s n before reading its accuracy.
      * @summary Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.
@@ -58,6 +83,14 @@ export declare const BenchmarkApiAxiosParamCreator: (configuration?: Configurati
      * @throws {RequiredError}
      */
     getBenchmarkPresets: (options?: RawAxiosRequestConfig) => Promise<RequestArgs>;
+    /**
+     * Records published claims: one to correct a number, many to import a leaderboard. Every row must carry a Source, because a claim without its citation is a number nobody can check — and an unattributed number in the published plane is indistinguishable from a measurement, which is the one confusion this whole surface is built to prevent.  Writes are append-only, so this never destroys the value it replaces. A vendor restating a score leaves both rows on disk, which is how the restating itself becomes visible.
+     * @summary Records published claims: one to correct a number, many to import a leaderboard.
+     * @param {PutClaimsIn} putClaimsIn
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    postBenchmarkClaims: (putClaimsIn: PutClaimsIn, options?: RawAxiosRequestConfig) => Promise<RequestArgs>;
     /**
      * Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.  It VALIDATES AND ECHOES: the definition is not persisted yet, so a preset accepted here is not one the model layer will resolve. Treat the response as a check on the blend, not a promise to serve it.  Defaults fill the shape rather than refusing it: an omitted rank becomes the arms in declared order and a panel below 1 becomes 1. The one real invariant is that rank may only name arms the blend declares — the same rule the model catalog enforces — and a rank naming anything else is a 422 listing exactly which entries were undeclared. A blend with no name or no arms is a 400.
      * @summary Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.
@@ -87,6 +120,18 @@ export declare const BenchmarkApiFp: (configuration?: Configuration) => {
      * @throws {RequiredError}
      */
     getBenchmarkCatalog(options?: RawAxiosRequestConfig): Promise<(axios?: AxiosInstance, basePath?: string) => AxiosPromise<BenchmarkCatalog>>;
+    /**
+     * Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered. It answers the operator\'s question — what does this arena currently believe someone else reported, and did we ship that or fix it.  Effective values only. The history of a key lives in the append-only file and is not what this op is for; a list that returned every superseded row would make the common question the hard one.
+     * @summary Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered.
+     * @param {string} [benchmark] Benchmark filters to one benchmark id. Empty returns every benchmark.
+     * @param {string} [model] Model filters to one model. Empty returns every model.
+     * @param {string} [provider] Provider filters to one lab or leaderboard — the way to read what a single source claims across every model it covers.
+     * @param {string} [source] Source filters to one citation, which is the finest grain there is: a source is what makes two claims about one model independent rather than a restatement of each other.
+     * @param {string} [protocol] Protocol filters by HOW a claim was scored, so provider cards can be read apart from third parties running their own harness.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    getBenchmarkClaims(benchmark?: string, model?: string, provider?: string, source?: string, protocol?: string, options?: RawAxiosRequestConfig): Promise<(axios?: AxiosInstance, basePath?: string) => AxiosPromise<ClaimsOut>>;
     /**
      * Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.  Pairing is what prevents the subset artifact — comparing one model\'s easy subset against another\'s full run — so n_common, not either arm\'s own coverage, is the number to read this by.  Both a and b are required. The benchmark defaults to gpqa_diamond.
      * @summary Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.
@@ -97,6 +142,15 @@ export declare const BenchmarkApiFp: (configuration?: Configuration) => {
      * @throws {RequiredError}
      */
     getBenchmarkCompare(a: string, b: string, benchmark?: string, options?: RawAxiosRequestConfig): Promise<(axios?: AxiosInstance, basePath?: string) => AxiosPromise<Pairing>>;
+    /**
+     * Returns each model\'s measured score per run over time, oldest first, with the change between runs.  This is the counterweight to a leaderboard: the board shows the latest run because that is what \"how good is it\" means, and a single latest number cannot distinguish a model that has always been strong from one that just improved, or from one that regressed after a provider changed something. Both matter for routing, and only one of them is visible on a board.  Runs with no id — attempts recorded before runs existed — group under the empty run, which is honestly what they are: one undated measurement.
+     * @summary Returns each model\'s measured score per run over time, oldest first, with the change between runs.
+     * @param {string} [benchmark] Benchmark is the catalog id to read, defaulting to gpqa_diamond.
+     * @param {string} [model] Model filters to one model. Empty returns every model measured.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    getBenchmarkHistory(benchmark?: string, model?: string, options?: RawAxiosRequestConfig): Promise<(axios?: AxiosInstance, basePath?: string) => AxiosPromise<HistoryOut>>;
     /**
      * Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.  The gap is the point of the arena; provider-reported claims have run materially hot against one standardized harness.  The two planes are NEVER blended, and that is the rule to read the rows by: a model we have measured but no vendor has claimed for shows published null, a model with only a claim shows measured null, and gap exists only where both do.  n is coverage and is not decoration: two measured numbers taken over different item counts are not comparable, so read the row\'s n before reading its accuracy.
      * @summary Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.
@@ -112,6 +166,14 @@ export declare const BenchmarkApiFp: (configuration?: Configuration) => {
      * @throws {RequiredError}
      */
     getBenchmarkPresets(options?: RawAxiosRequestConfig): Promise<(axios?: AxiosInstance, basePath?: string) => AxiosPromise<PresetList>>;
+    /**
+     * Records published claims: one to correct a number, many to import a leaderboard. Every row must carry a Source, because a claim without its citation is a number nobody can check — and an unattributed number in the published plane is indistinguishable from a measurement, which is the one confusion this whole surface is built to prevent.  Writes are append-only, so this never destroys the value it replaces. A vendor restating a score leaves both rows on disk, which is how the restating itself becomes visible.
+     * @summary Records published claims: one to correct a number, many to import a leaderboard.
+     * @param {PutClaimsIn} putClaimsIn
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    postBenchmarkClaims(putClaimsIn: PutClaimsIn, options?: RawAxiosRequestConfig): Promise<(axios?: AxiosInstance, basePath?: string) => AxiosPromise<PutClaimsOut>>;
     /**
      * Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.  It VALIDATES AND ECHOES: the definition is not persisted yet, so a preset accepted here is not one the model layer will resolve. Treat the response as a check on the blend, not a promise to serve it.  Defaults fill the shape rather than refusing it: an omitted rank becomes the arms in declared order and a panel below 1 becomes 1. The one real invariant is that rank may only name arms the blend declares — the same rule the model catalog enforces — and a rank naming anything else is a 422 listing exactly which entries were undeclared. A blend with no name or no arms is a 400.
      * @summary Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.
@@ -141,6 +203,14 @@ export declare const BenchmarkApiFactory: (configuration?: Configuration, basePa
      * @throws {RequiredError}
      */
     getBenchmarkCatalog(options?: RawAxiosRequestConfig): AxiosPromise<BenchmarkCatalog>;
+    /**
+     * Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered. It answers the operator\'s question — what does this arena currently believe someone else reported, and did we ship that or fix it.  Effective values only. The history of a key lives in the append-only file and is not what this op is for; a list that returned every superseded row would make the common question the hard one.
+     * @summary Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered.
+     * @param {BenchmarkApiGetBenchmarkClaimsRequest} requestParameters Request parameters.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    getBenchmarkClaims(requestParameters?: BenchmarkApiGetBenchmarkClaimsRequest, options?: RawAxiosRequestConfig): AxiosPromise<ClaimsOut>;
     /**
      * Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.  Pairing is what prevents the subset artifact — comparing one model\'s easy subset against another\'s full run — so n_common, not either arm\'s own coverage, is the number to read this by.  Both a and b are required. The benchmark defaults to gpqa_diamond.
      * @summary Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.
@@ -149,6 +219,14 @@ export declare const BenchmarkApiFactory: (configuration?: Configuration, basePa
      * @throws {RequiredError}
      */
     getBenchmarkCompare(requestParameters: BenchmarkApiGetBenchmarkCompareRequest, options?: RawAxiosRequestConfig): AxiosPromise<Pairing>;
+    /**
+     * Returns each model\'s measured score per run over time, oldest first, with the change between runs.  This is the counterweight to a leaderboard: the board shows the latest run because that is what \"how good is it\" means, and a single latest number cannot distinguish a model that has always been strong from one that just improved, or from one that regressed after a provider changed something. Both matter for routing, and only one of them is visible on a board.  Runs with no id — attempts recorded before runs existed — group under the empty run, which is honestly what they are: one undated measurement.
+     * @summary Returns each model\'s measured score per run over time, oldest first, with the change between runs.
+     * @param {BenchmarkApiGetBenchmarkHistoryRequest} requestParameters Request parameters.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    getBenchmarkHistory(requestParameters?: BenchmarkApiGetBenchmarkHistoryRequest, options?: RawAxiosRequestConfig): AxiosPromise<HistoryOut>;
     /**
      * Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.  The gap is the point of the arena; provider-reported claims have run materially hot against one standardized harness.  The two planes are NEVER blended, and that is the rule to read the rows by: a model we have measured but no vendor has claimed for shows published null, a model with only a claim shows measured null, and gap exists only where both do.  n is coverage and is not decoration: two measured numbers taken over different item counts are not comparable, so read the row\'s n before reading its accuracy.
      * @summary Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.
@@ -164,6 +242,14 @@ export declare const BenchmarkApiFactory: (configuration?: Configuration, basePa
      * @throws {RequiredError}
      */
     getBenchmarkPresets(options?: RawAxiosRequestConfig): AxiosPromise<PresetList>;
+    /**
+     * Records published claims: one to correct a number, many to import a leaderboard. Every row must carry a Source, because a claim without its citation is a number nobody can check — and an unattributed number in the published plane is indistinguishable from a measurement, which is the one confusion this whole surface is built to prevent.  Writes are append-only, so this never destroys the value it replaces. A vendor restating a score leaves both rows on disk, which is how the restating itself becomes visible.
+     * @summary Records published claims: one to correct a number, many to import a leaderboard.
+     * @param {BenchmarkApiPostBenchmarkClaimsRequest} requestParameters Request parameters.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     */
+    postBenchmarkClaims(requestParameters: BenchmarkApiPostBenchmarkClaimsRequest, options?: RawAxiosRequestConfig): AxiosPromise<PutClaimsOut>;
     /**
      * Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.  It VALIDATES AND ECHOES: the definition is not persisted yet, so a preset accepted here is not one the model layer will resolve. Treat the response as a check on the blend, not a promise to serve it.  Defaults fill the shape rather than refusing it: an omitted rank becomes the arms in declared order and a panel below 1 becomes 1. The one real invariant is that rank may only name arms the blend declares — the same rule the model catalog enforces — and a rank naming anything else is a 422 listing exactly which entries were undeclared. A blend with no name or no arms is a 400.
      * @summary Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.
@@ -181,6 +267,43 @@ export declare const BenchmarkApiFactory: (configuration?: Configuration, basePa
      */
     postBenchmarkRuns(requestParameters: BenchmarkApiPostBenchmarkRunsRequest, options?: RawAxiosRequestConfig): AxiosPromise<Admission>;
 };
+/**
+ * Request parameters for getBenchmarkClaims operation in BenchmarkApi.
+ * @export
+ * @interface BenchmarkApiGetBenchmarkClaimsRequest
+ */
+export interface BenchmarkApiGetBenchmarkClaimsRequest {
+    /**
+     * Benchmark filters to one benchmark id. Empty returns every benchmark.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkClaims
+     */
+    readonly benchmark?: string;
+    /**
+     * Model filters to one model. Empty returns every model.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkClaims
+     */
+    readonly model?: string;
+    /**
+     * Provider filters to one lab or leaderboard — the way to read what a single source claims across every model it covers.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkClaims
+     */
+    readonly provider?: string;
+    /**
+     * Source filters to one citation, which is the finest grain there is: a source is what makes two claims about one model independent rather than a restatement of each other.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkClaims
+     */
+    readonly source?: string;
+    /**
+     * Protocol filters by HOW a claim was scored, so provider cards can be read apart from third parties running their own harness.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkClaims
+     */
+    readonly protocol?: string;
+}
 /**
  * Request parameters for getBenchmarkCompare operation in BenchmarkApi.
  * @export
@@ -206,6 +329,25 @@ export interface BenchmarkApiGetBenchmarkCompareRequest {
      */
     readonly benchmark?: string;
 }
+/**
+ * Request parameters for getBenchmarkHistory operation in BenchmarkApi.
+ * @export
+ * @interface BenchmarkApiGetBenchmarkHistoryRequest
+ */
+export interface BenchmarkApiGetBenchmarkHistoryRequest {
+    /**
+     * Benchmark is the catalog id to read, defaulting to gpqa_diamond.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkHistory
+     */
+    readonly benchmark?: string;
+    /**
+     * Model filters to one model. Empty returns every model measured.
+     * @type {string}
+     * @memberof BenchmarkApiGetBenchmarkHistory
+     */
+    readonly model?: string;
+}
 /**
  * Request parameters for getBenchmarkLeaderboard operation in BenchmarkApi.
  * @export
@@ -219,6 +361,19 @@ export interface BenchmarkApiGetBenchmarkLeaderboardRequest {
      */
     readonly benchmark?: string;
 }
+/**
+ * Request parameters for postBenchmarkClaims operation in BenchmarkApi.
+ * @export
+ * @interface BenchmarkApiPostBenchmarkClaimsRequest
+ */
+export interface BenchmarkApiPostBenchmarkClaimsRequest {
+    /**
+     *
+     * @type {PutClaimsIn}
+     * @memberof BenchmarkApiPostBenchmarkClaims
+     */
+    readonly putClaimsIn: PutClaimsIn;
+}
 /**
  * Request parameters for postBenchmarkPresets operation in BenchmarkApi.
  * @export
@@ -260,6 +415,15 @@ export declare class BenchmarkApi extends BaseAPI {
      * @memberof BenchmarkApi
      */
     getBenchmarkCatalog(options?: RawAxiosRequestConfig): Promise<globalAxios.AxiosResponse<BenchmarkCatalog, any, {}>>;
+    /**
+     * Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered. It answers the operator\'s question — what does this arena currently believe someone else reported, and did we ship that or fix it.  Effective values only. The history of a key lives in the append-only file and is not what this op is for; a list that returned every superseded row would make the common question the hard one.
+     * @summary Lists the effective published claims: what the leaderboard will use for each (benchmark, model) after the seed, the import and any stored correction are layered.
+     * @param {BenchmarkApiGetBenchmarkClaimsRequest} requestParameters Request parameters.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     * @memberof BenchmarkApi
+     */
+    getBenchmarkClaims(requestParameters?: BenchmarkApiGetBenchmarkClaimsRequest, options?: RawAxiosRequestConfig): Promise<globalAxios.AxiosResponse<ClaimsOut, any, {}>>;
     /**
      * Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.  Pairing is what prevents the subset artifact — comparing one model\'s easy subset against another\'s full run — so n_common, not either arm\'s own coverage, is the number to read this by.  Both a and b are required. The benchmark defaults to gpqa_diamond.
      * @summary Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p.
@@ -269,6 +433,15 @@ export declare class BenchmarkApi extends BaseAPI {
      * @memberof BenchmarkApi
      */
     getBenchmarkCompare(requestParameters: BenchmarkApiGetBenchmarkCompareRequest, options?: RawAxiosRequestConfig): Promise<globalAxios.AxiosResponse<Pairing, any, {}>>;
+    /**
+     * Returns each model\'s measured score per run over time, oldest first, with the change between runs.  This is the counterweight to a leaderboard: the board shows the latest run because that is what \"how good is it\" means, and a single latest number cannot distinguish a model that has always been strong from one that just improved, or from one that regressed after a provider changed something. Both matter for routing, and only one of them is visible on a board.  Runs with no id — attempts recorded before runs existed — group under the empty run, which is honestly what they are: one undated measurement.
+     * @summary Returns each model\'s measured score per run over time, oldest first, with the change between runs.
+     * @param {BenchmarkApiGetBenchmarkHistoryRequest} requestParameters Request parameters.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     * @memberof BenchmarkApi
+     */
+    getBenchmarkHistory(requestParameters?: BenchmarkApiGetBenchmarkHistoryRequest, options?: RawAxiosRequestConfig): Promise<globalAxios.AxiosResponse<HistoryOut, any, {}>>;
     /**
      * Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.  The gap is the point of the arena; provider-reported claims have run materially hot against one standardized harness.  The two planes are NEVER blended, and that is the rule to read the rows by: a model we have measured but no vendor has claimed for shows published null, a model with only a claim shows measured null, and gap exists only where both do.  n is coverage and is not decoration: two measured numbers taken over different item counts are not comparable, so read the row\'s n before reading its accuracy.
      * @summary Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them.
@@ -286,6 +459,15 @@ export declare class BenchmarkApi extends BaseAPI {
      * @memberof BenchmarkApi
      */
     getBenchmarkPresets(options?: RawAxiosRequestConfig): Promise<globalAxios.AxiosResponse<PresetList, any, {}>>;
+    /**
+     * Records published claims: one to correct a number, many to import a leaderboard. Every row must carry a Source, because a claim without its citation is a number nobody can check — and an unattributed number in the published plane is indistinguishable from a measurement, which is the one confusion this whole surface is built to prevent.  Writes are append-only, so this never destroys the value it replaces. A vendor restating a score leaves both rows on disk, which is how the restating itself becomes visible.
+     * @summary Records published claims: one to correct a number, many to import a leaderboard.
+     * @param {BenchmarkApiPostBenchmarkClaimsRequest} requestParameters Request parameters.
+     * @param {*} [options] Override http request option.
+     * @throws {RequiredError}
+     * @memberof BenchmarkApi
+     */
+    postBenchmarkClaims(requestParameters: BenchmarkApiPostBenchmarkClaimsRequest, options?: RawAxiosRequestConfig): Promise<globalAxios.AxiosResponse<PutClaimsOut, any, {}>>;
     /**
      * Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.  It VALIDATES AND ECHOES: the definition is not persisted yet, so a preset accepted here is not one the model layer will resolve. Treat the response as a check on the blend, not a promise to serve it.  Defaults fill the shape rather than refusing it: an omitted rank becomes the arms in declared order and a panel below 1 becomes 1. The one real invariant is that rank may only name arms the blend declares — the same rule the model catalog enforces — and a rank naming anything else is a 422 listing exactly which entries were undeclared. A blend with no name or no arms is a 400.
      * @summary Validates a router blend — its name, its arms, the rank they escalate through and the panel fan-out width — and answers 202 with the preset and the enso-<name> it would be served as.