llama-stack-client 0.6.1 → 0.7.1

package/dist/version.js

@@ -6,5 +6,5 @@
 // the root directory of this source tree.
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.VERSION = void 0;
-exports.VERSION = '0.6.1'; // x-release-please-version
+exports.VERSION = '0.7.1'; // x-release-please-version
 //# sourceMappingURL=version.js.map

package/dist/version.mjs

@@ -3,5 +3,5 @@
 //
 // This source code is licensed under the terms described in the LICENSE file in
 // the root directory of this source tree.
-export const VERSION = '0.6.1'; // x-release-please-version
+export const VERSION = '0.7.1'; // x-release-please-version
 //# sourceMappingURL=version.mjs.map

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "llama-stack-client",
-  "version": "0.6.1",
+  "version": "0.7.1",
   "description": "The official TypeScript library for the Llama Stack Client API",
   "author": "Llama Stack Client <llamastack@meta.com>",
   "types": "dist/index.d.ts",

package/scripts/mock

@@ -19,23 +19,34 @@ fi
 
 echo "==> Starting mock server with URL ${URL}"
 
-# Run prism mock on the given spec
+# Run steady mock on the given spec
 if [ "$1" == "--daemon" ]; then
-  npm exec --package=@stainless-api/prism-cli@5.15.0 -- prism mock "$URL" &> .prism.log &
+  # Pre-install the package so the download doesn't eat into the startup timeout
+  npm exec --package=@stdy/cli@0.20.2 -- steady --version
 
-  # Wait for server to come online
+  npm exec --package=@stdy/cli@0.20.2 -- steady --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets "$URL" &> .stdy.log &
+
+  # Wait for server to come online via health endpoint (max 30s)
   echo -n "Waiting for server"
-  while ! grep -q "✖  fatal\|Prism is listening" ".prism.log" ; do
+  attempts=0
+  while ! curl --silent --fail "http://127.0.0.1:4010/_x-steady/health" >/dev/null 2>&1; do
+    if ! kill -0 $! 2>/dev/null; then
+      echo
+      cat .stdy.log
+      exit 1
+    fi
+    attempts=$((attempts + 1))
+    if [ "$attempts" -ge 300 ]; then
+      echo
+      echo "Timed out waiting for Steady server to start"
+      cat .stdy.log
+      exit 1
+    fi
     echo -n "."
     sleep 0.1
   done
 
-  if grep -q "✖  fatal" ".prism.log"; then
-    cat .prism.log
-    exit 1
-  fi
-
   echo
 else
-  npm exec --package=@stainless-api/prism-cli@5.15.0 -- prism mock "$URL"
+  npm exec --package=@stdy/cli@0.20.2 -- steady --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets "$URL"
 fi

package/scripts/test

@@ -9,8 +9,8 @@ GREEN='\033[0;32m'
 YELLOW='\033[0;33m'
 NC='\033[0m' # No Color
 
-function prism_is_running() {
-  curl --silent "http://localhost:4010" >/dev/null 2>&1
+function steady_is_running() {
+  curl --silent "http://127.0.0.1:4010/_x-steady/health" >/dev/null 2>&1
 }
 
 kill_server_on_port() {
@@ -25,7 +25,7 @@ function is_overriding_api_base_url() {
   [ -n "$TEST_API_BASE_URL" ]
 }
 
-if ! is_overriding_api_base_url && ! prism_is_running ; then
+if ! is_overriding_api_base_url && ! steady_is_running ; then
   # When we exit this script, make sure to kill the background mock server process
   trap 'kill_server_on_port 4010' EXIT
 
@@ -36,19 +36,19 @@ fi
 if is_overriding_api_base_url ; then
   echo -e "${GREEN}✔ Running tests against ${TEST_API_BASE_URL}${NC}"
   echo
-elif ! prism_is_running ; then
-  echo -e "${RED}ERROR:${NC} The test suite will not run without a mock Prism server"
+elif ! steady_is_running ; then
+  echo -e "${RED}ERROR:${NC} The test suite will not run without a mock Steady server"
   echo -e "running against your OpenAPI spec."
   echo
   echo -e "To run the server, pass in the path or url of your OpenAPI"
-  echo -e "spec to the prism command:"
+  echo -e "spec to the steady command:"
   echo
-  echo -e "  \$ ${YELLOW}npm exec --package=@stainless-api/prism-cli@5.15.0 -- prism mock path/to/your.openapi.yml${NC}"
+  echo -e "  \$ ${YELLOW}npm exec --package=@stdy/cli@0.20.2 -- steady path/to/your.openapi.yml --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets${NC}"
   echo
 
   exit 1
 else
-  echo -e "${GREEN}✔ Mock prism server is running with your OpenAPI spec${NC}"
+  echo -e "${GREEN}✔ Mock steady server is running with your OpenAPI spec${NC}"
   echo
 fi
 

package/src/core.ts

@@ -13,6 +13,7 @@ import {
   APIConnectionTimeoutError,
   APIUserAbortError,
 } from './error';
+import { stringifyQuery } from './internal/utils/query';
 import {
   kind as shimsKind,
   type Readable,
@@ -543,32 +544,20 @@ export abstract class APIClient {
       : new URL(baseURL + (baseURL.endsWith('/') && path.startsWith('/') ? path.slice(1) : path));
 
     const defaultQuery = this.defaultQuery();
-    if (!isEmptyObj(defaultQuery)) {
-      query = { ...defaultQuery, ...query } as Req;
+    const pathQuery = Object.fromEntries(url.searchParams);
+    if (!isEmptyObj(defaultQuery) || !isEmptyObj(pathQuery)) {
+      query = { ...pathQuery, ...defaultQuery, ...query } as Req;
     }
 
     if (typeof query === 'object' && query && !Array.isArray(query)) {
-      url.search = this.stringifyQuery(query as Record<string, unknown>);
+      url.search = this.stringifyQuery(query);
     }
 
     return url.toString();
   }
 
-  protected stringifyQuery(query: Record<string, unknown>): string {
-    return Object.entries(query)
-      .filter(([_, value]) => typeof value !== 'undefined')
-      .map(([key, value]) => {
-        if (typeof value === 'string' || typeof value === 'number' || typeof value === 'boolean') {
-          return `${encodeURIComponent(key)}=${encodeURIComponent(value)}`;
-        }
-        if (value === null) {
-          return `${encodeURIComponent(key)}=`;
-        }
-        throw new LlamaStackClientError(
-          `Cannot stringify type ${typeof value}; Expected string, number, boolean, or null. If you need to pass nested query parameters, you can manually encode them, e.g. { query: { 'foo[key1]': value1, 'foo[key2]': value2 } }, and please open a GitHub issue requesting better support for your use case.`,
-        );
-      })
-      .join('&');
+  protected stringifyQuery(query: object | Record<string, unknown>): string {
+    return stringifyQuery(query);
   }
 
   async fetchWithTimeout(
@@ -650,9 +639,9 @@ export abstract class APIClient {
       }
     }
 
-    // If the API asks us to wait a certain amount of time (and it's a reasonable amount),
-    // just do what it says, but otherwise calculate a default
-    if (!(timeoutMillis && 0 <= timeoutMillis && timeoutMillis < 60 * 1000)) {
+    // If the API asks us to wait a certain amount of time, do what it says.
+    // Otherwise calculate a default.
+    if (timeoutMillis === undefined) {
       const maxRetries = options.maxRetries ?? this.maxRetries;
       timeoutMillis = this.calculateDefaultRetryTimeoutMillis(retriesRemaining, maxRetries);
     }

package/src/index.ts

@@ -8,7 +8,7 @@
 // File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 import { type Agent } from './_shims/index';
-import * as qs from './internal/qs';
+import { stringifyQuery } from './internal/utils/query';
 import * as Core from './core';
 import * as Errors from './error';
 import * as Pagination from './pagination';
@@ -75,22 +75,6 @@ import {
   ShieldRegisterParams,
   Shields,
 } from './resources/shields';
-import {
-  ToolDef,
-  ToolInvocationResult,
-  ToolRuntime,
-  ToolRuntimeInvokeToolParams,
-  ToolRuntimeListToolsParams,
-  ToolRuntimeListToolsResponse,
-} from './resources/tool-runtime';
-import {
-  ListToolGroupsResponse,
-  ToolGroup,
-  ToolgroupListResponse,
-  ToolgroupRegisterParams,
-  Toolgroups,
-} from './resources/toolgroups';
-import { ToolListParams, ToolListResponse, Tools } from './resources/tools';
 import {
   QueryChunksResponse,
   VectorIo,
@@ -111,8 +95,6 @@ import {
   ListModelsResponse,
   Model,
   ModelListResponse,
-  ModelRegisterParams,
-  ModelRegisterResponse,
   ModelRetrieveResponse,
   Models,
 } from './resources/models/models';
@@ -261,27 +243,72 @@ export class LlamaStackClient extends Core.APIClient {
     this.apiKey = apiKey;
   }
 
-  toolgroups: API.Toolgroups = new API.Toolgroups(this);
-  tools: API.Tools = new API.Tools(this);
-  toolRuntime: API.ToolRuntime = new API.ToolRuntime(this);
   responses: API.Responses = new API.Responses(this);
+  /**
+   * Protocol for prompt management operations.
+   */
   prompts: API.Prompts = new API.Prompts(this);
+  /**
+   * Protocol for conversation management operations.
+   */
   conversations: API.Conversations = new API.Conversations(this);
+  /**
+   * APIs for inspecting the Llama Stack service, including health status, available API routes with methods and implementing providers.
+   */
   inspect: API.Inspect = new API.Inspect(this);
+  /**
+   * Llama Stack Inference API for generating completions, chat completions, and embeddings.
+   *
+   * This API provides the raw interface to the underlying models. Three kinds of models are supported:
+   * - LLM models: these models generate "raw" and "chat" (conversational) completions.
+   * - Embedding models: these models generate embeddings to be used for semantic search.
+   * - Rerank models: these models reorder the documents based on their relevance to a query.
+   */
   embeddings: API.Embeddings = new API.Embeddings(this);
   chat: API.Chat = new API.Chat(this);
+  /**
+   * Llama Stack Inference API for generating completions, chat completions, and embeddings.
+   *
+   * This API provides the raw interface to the underlying models. Three kinds of models are supported:
+   * - LLM models: these models generate "raw" and "chat" (conversational) completions.
+   * - Embedding models: these models generate embeddings to be used for semantic search.
+   * - Rerank models: these models reorder the documents based on their relevance to a query.
+   */
   completions: API.Completions = new API.Completions(this);
   vectorIo: API.VectorIo = new API.VectorIo(this);
   vectorStores: API.VectorStores = new API.VectorStores(this);
   models: API.Models = new API.Models(this);
+  /**
+   * Providers API for inspecting, listing, and modifying providers and their configurations.
+   */
   providers: API.Providers = new API.Providers(this);
+  /**
+   * APIs for inspecting the Llama Stack service, including health status, available API routes with methods and implementing providers.
+   */
   routes: API.Routes = new API.Routes(this);
+  /**
+   * OpenAI-compatible Moderations API.
+   */
   moderations: API.Moderations = new API.Moderations(this);
+  /**
+   * OpenAI-compatible Moderations API.
+   */
   safety: API.Safety = new API.Safety(this);
   shields: API.Shields = new API.Shields(this);
   scoring: API.Scoring = new API.Scoring(this);
   scoringFunctions: API.ScoringFunctions = new API.ScoringFunctions(this);
+  /**
+   * This API is used to upload documents that can be used with other Llama Stack APIs.
+   */
   files: API.Files = new API.Files(this);
+  /**
+   * The API is designed to allow use of openai client libraries for seamless integration.
+   *
+   * This API provides the following extensions:
+   *  - idempotent batch creation
+   *
+   * Note: This API is currently under active development and may undergo changes.
+   */
   batches: API.Batches = new API.Batches(this);
   alpha: API.Alpha = new API.Alpha(this);
   beta: API.Beta = new API.Beta(this);
@@ -311,8 +338,8 @@ export class LlamaStackClient extends Core.APIClient {
     return { Authorization: `Bearer ${this.apiKey}` };
   }
 
-  protected override stringifyQuery(query: Record<string, unknown>): string {
-    return qs.stringify(query, { arrayFormat: 'comma' });
+  protected override stringifyQuery(query: object | Record<string, unknown>): string {
+    return stringifyQuery(query);
   }
 
   static LlamaStackClient = this;
@@ -336,9 +363,6 @@ export class LlamaStackClient extends Core.APIClient {
   static fileFromPath = Uploads.fileFromPath;
 }
 
-LlamaStackClient.Toolgroups = Toolgroups;
-LlamaStackClient.Tools = Tools;
-LlamaStackClient.ToolRuntime = ToolRuntime;
 LlamaStackClient.Responses = Responses;
 LlamaStackClient.ResponseListResponsesOpenAICursorPage = ResponseListResponsesOpenAICursorPage;
 LlamaStackClient.Prompts = Prompts;
@@ -380,25 +404,6 @@ export declare namespace LlamaStackClient {
     type OpenAICursorPageResponse as OpenAICursorPageResponse,
   };
 
-  export {
-    Toolgroups as Toolgroups,
-    type ListToolGroupsResponse as ListToolGroupsResponse,
-    type ToolGroup as ToolGroup,
-    type ToolgroupListResponse as ToolgroupListResponse,
-    type ToolgroupRegisterParams as ToolgroupRegisterParams,
-  };
-
-  export { Tools as Tools, type ToolListResponse as ToolListResponse, type ToolListParams as ToolListParams };
-
-  export {
-    ToolRuntime as ToolRuntime,
-    type ToolDef as ToolDef,
-    type ToolInvocationResult as ToolInvocationResult,
-    type ToolRuntimeListToolsResponse as ToolRuntimeListToolsResponse,
-    type ToolRuntimeInvokeToolParams as ToolRuntimeInvokeToolParams,
-    type ToolRuntimeListToolsParams as ToolRuntimeListToolsParams,
-  };
-
   export {
     Responses as Responses,
     type ResponseObject as ResponseObject,
@@ -475,8 +480,6 @@ export declare namespace LlamaStackClient {
     type Model as Model,
     type ModelRetrieveResponse as ModelRetrieveResponse,
     type ModelListResponse as ModelListResponse,
-    type ModelRegisterResponse as ModelRegisterResponse,
-    type ModelRegisterParams as ModelRegisterParams,
   };
 
   export { Providers as Providers, type ProviderListResponse as ProviderListResponse };

package/{dist/src/resources/alpha/post-training.ts → src/internal/utils/query.ts}

@@ -6,4 +6,8 @@
 //
 // File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
-export * from './post-training/index';
+import * as qs from '../qs/stringify';
+
+export function stringifyQuery(query: object | Record<string, unknown>) {
+  return qs.stringify(query, { arrayFormat: 'comma' });
+}

package/src/resources/alpha/alpha.ts

@@ -30,40 +30,20 @@ import {
   EvaluateResponse,
   Job,
 } from './eval/eval';
-import * as PostTrainingAPI from './post-training/post-training';
-import {
-  AlgorithmConfig,
-  ListPostTrainingJobsResponse,
-  PostTraining,
-  PostTrainingJob,
-  PostTrainingPreferenceOptimizeParams,
-  PostTrainingSupervisedFineTuneParams,
-} from './post-training/post-training';
 
 export class Alpha extends APIResource {
-  postTraining: PostTrainingAPI.PostTraining = new PostTrainingAPI.PostTraining(this._client);
   benchmarks: BenchmarksAPI.Benchmarks = new BenchmarksAPI.Benchmarks(this._client);
   eval: EvalAPI.Eval = new EvalAPI.Eval(this._client);
   admin: AdminAPI.Admin = new AdminAPI.Admin(this._client);
   inference: InferenceAPI.Inference = new InferenceAPI.Inference(this._client);
 }
 
-Alpha.PostTraining = PostTraining;
 Alpha.Benchmarks = Benchmarks;
 Alpha.Eval = Eval;
 Alpha.Admin = Admin;
 Alpha.Inference = Inference;
 
 export declare namespace Alpha {
-  export {
-    PostTraining as PostTraining,
-    type AlgorithmConfig as AlgorithmConfig,
-    type ListPostTrainingJobsResponse as ListPostTrainingJobsResponse,
-    type PostTrainingJob as PostTrainingJob,
-    type PostTrainingPreferenceOptimizeParams as PostTrainingPreferenceOptimizeParams,
-    type PostTrainingSupervisedFineTuneParams as PostTrainingSupervisedFineTuneParams,
-  };
-
   export {
     Benchmarks as Benchmarks,
     type Benchmark as Benchmark,

package/src/resources/alpha/eval/eval.ts

@@ -12,6 +12,9 @@ import * as Shared from '../../shared';
 import * as JobsAPI from './jobs';
 import { Jobs } from './jobs';
 
+/**
+ * Llama Stack Evaluation API for running evaluations on model and agent candidates.
+ */
 export class Eval extends APIResource {
   jobs: JobsAPI.Jobs = new JobsAPI.Jobs(this._client);
 

package/src/resources/alpha/eval/jobs.ts

@@ -10,6 +10,9 @@ import { APIResource } from '../../../resource';
 import * as Core from '../../../core';
 import * as EvalAPI from './eval';
 
+/**
+ * Llama Stack Evaluation API for running evaluations on model and agent candidates.
+ */
 export class Jobs extends APIResource {
   /**
    * Get the result of a job.

package/src/resources/alpha/index.ts

@@ -26,11 +26,3 @@ export {
   type EvalRunEvalAlphaParams,
 } from './eval/index';
 export { Inference, type InferenceRerankResponse, type InferenceRerankParams } from './inference';
-export {
-  PostTraining,
-  type AlgorithmConfig,
-  type ListPostTrainingJobsResponse,
-  type PostTrainingJob,
-  type PostTrainingPreferenceOptimizeParams,
-  type PostTrainingSupervisedFineTuneParams,
-} from './post-training/index';

package/src/resources/alpha/inference.ts

@@ -9,6 +9,14 @@
 import { APIResource } from '../../resource';
 import * as Core from '../../core';
 
+/**
+ * Llama Stack Inference API for generating completions, chat completions, and embeddings.
+ *
+ * This API provides the raw interface to the underlying models. Three kinds of models are supported:
+ * - LLM models: these models generate "raw" and "chat" (conversational) completions.
+ * - Embedding models: these models generate embeddings to be used for semantic search.
+ * - Rerank models: these models reorder the documents based on their relevance to a query.
+ */
 export class Inference extends APIResource {
   /**
    * Rerank a list of documents based on their relevance to a query.

package/src/resources/batches.ts

@@ -11,6 +11,14 @@ import { isRequestOptions } from '../core';
 import * as Core from '../core';
 import { OpenAICursorPage, type OpenAICursorPageParams } from '../pagination';
 
+/**
+ * The API is designed to allow use of openai client libraries for seamless integration.
+ *
+ * This API provides the following extensions:
+ *  - idempotent batch creation
+ *
+ * Note: This API is currently under active development and may undergo changes.
+ */
 export class Batches extends APIResource {
   /**
    * Create a new batch for processing multiple API requests.

package/src/resources/beta/datasets.ts

@@ -124,7 +124,7 @@ export interface DatasetRetrieveResponse {
   /**
    * Purpose of the dataset indicating its intended use
    */
-  purpose: 'post-training/messages' | 'eval/question-answer' | 'eval/messages-answer';
+  purpose: 'eval/question-answer' | 'eval/messages-answer';
 
   /**
    * Data source configuration for the dataset
@@ -205,7 +205,7 @@ export namespace DatasetListResponse {
     /**
      * Purpose of the dataset indicating its intended use
      */
-    purpose: 'post-training/messages' | 'eval/question-answer' | 'eval/messages-answer';
+    purpose: 'eval/question-answer' | 'eval/messages-answer';
 
     /**
      * Data source configuration for the dataset
@@ -292,7 +292,7 @@ export interface DatasetRegisterResponse {
   /**
    * Purpose of the dataset indicating its intended use
    */
-  purpose: 'post-training/messages' | 'eval/question-answer' | 'eval/messages-answer';
+  purpose: 'eval/question-answer' | 'eval/messages-answer';
 
   /**
    * Data source configuration for the dataset
@@ -351,11 +351,6 @@ export namespace DatasetRegisterResponse {
 }
 
 export interface DatasetAppendrowsParams {
-  /**
-   * The ID of the dataset to append the rows to.
-   */
-  body_dataset_id: string;
-
   /**
    * The rows to append to the dataset.
    */
@@ -378,7 +373,7 @@ export interface DatasetRegisterParams {
   /**
    * The purpose of the dataset.
    */
-  purpose: 'post-training/messages' | 'eval/question-answer' | 'eval/messages-answer';
+  purpose: 'eval/question-answer' | 'eval/messages-answer';
 
   /**
    * The data source of the dataset.