@promptbook/openai 0.89.0-9 → 0.92.0-10

package/README.md

@@ -447,6 +447,10 @@ But unlike programming languages, it is designed to be understandable by non-pro
 
 
 
+## 🔒 Security
+
+For information on reporting security vulnerabilities, see our [Security Policy](./SECURITY.md).
+
 ## 📦 Packages _(for developers)_
 
 This library is divided into several packages, all are published from [single monorepo](https://github.com/webgptorg/promptbook).
@@ -503,7 +507,7 @@ The following glossary is used to clarify certain concepts:
 ### General LLM / AI terms
 
 -   **Prompt drift** is a phenomenon where the AI model starts to generate outputs that are not aligned with the original prompt. This can happen due to the model's training data, the prompt's wording, or the model's architecture.
--   **Pipeline, workflow or chain** is a sequence of tasks that are executed in a specific order. In the context of AI, a pipeline can refer to a sequence of AI models that are used to process data.
+-   [**Pipeline, workflow scenario or chain** is a sequence of tasks that are executed in a specific order. In the context of AI, a pipeline can refer to a sequence of AI models that are used to process data.](https://github.com/webgptorg/promptbook/discussions/88)
 -   **Fine-tuning** is a process where a pre-trained AI model is further trained on a specific dataset to improve its performance on a specific task.
 -   **Zero-shot learning** is a machine learning paradigm where a model is trained to perform a task without any labeled examples. Instead, the model is provided with a description of the task and is expected to generate the correct output.
 -   **Few-shot learning** is a machine learning paradigm where a model is trained to perform a task with only a few labeled examples. This is in contrast to traditional machine learning, where models are trained on large datasets.
@@ -511,10 +515,6 @@ The following glossary is used to clarify certain concepts:
 -   **Retrieval-augmented generation** is a machine learning paradigm where a model generates text by retrieving relevant information from a large database of text. This approach combines the benefits of generative models and retrieval models.
 -   **Longtail** refers to non-common or rare events, items, or entities that are not well-represented in the training data of machine learning models. Longtail items are often challenging for models to predict accurately.
 
-
-
-
-
 _Note: This section is not complete dictionary, more list of general AI / LLM terms that has connection with Promptbook_
 
 
@@ -628,6 +628,8 @@ See [TODO.md](./TODO.md)
 
 ## 🖋️ Contributing
 
-We are open to pull requests, feedback, and suggestions.
+You can also ⭐ star the project, [follow us on GitHub](https://github.com/hejny) or [various other social networks](https://www.pavolhejny.com/contact/).We are open to [pull requests, feedback, and suggestions](./CONTRIBUTING.md).
+
+## 📞 Support
 
-You can also ⭐ star the project, [follow us on GitHub](https://github.com/hejny) or [various other social networks](https://www.pavolhejny.com/contact/).
+If you need help or have questions, please check our [Support Resources](./SUPPORT.md).

package/esm/index.es.js

@@ -1,6 +1,7 @@
 import colors from 'colors';
 import spaceTrim$1, { spaceTrim } from 'spacetrim';
 import { randomBytes } from 'crypto';
+import Bottleneck from 'bottleneck';
 import OpenAI from 'openai';
 
 // ⚠️ WARNING: This code has been generated so that any manual changes will be overwritten
@@ -17,7 +18,7 @@ const BOOK_LANGUAGE_VERSION = '1.0.0';
  * @generated
  * @see https://github.com/webgptorg/promptbook
  */
-const PROMPTBOOK_ENGINE_VERSION = '0.89.0-9';
+const PROMPTBOOK_ENGINE_VERSION = '0.92.0-10';
 /**
  * TODO: string_promptbook_version should be constrained to the all versions of Promptbook engine
  * Note: [💞] Ignore a discrepancy between file name and entity name
@@ -117,7 +118,7 @@ class PipelineExecutionError extends Error {
     }
 }
 /**
- * TODO: !!!!!! Add id to all errors
+ * TODO: [🧠][🌂] Add id to all errors
  */
 
 /**
@@ -267,6 +268,7 @@ const VALUE_STRINGS = {
     infinity: '(infinity; ∞)',
     negativeInfinity: '(negative infinity; -∞)',
     unserializable: '(unserializable value)',
+    circular: '(circular JSON)',
 };
 /**
  * Small number limit
@@ -374,6 +376,54 @@ class UnexpectedError extends Error {
     }
 }
 
+/**
+ * This error type indicates that somewhere in the code non-Error object was thrown and it was wrapped into the `WrappedError`
+ *
+ * @public exported from `@promptbook/core`
+ */
+class WrappedError extends Error {
+    constructor(whatWasThrown) {
+        const tag = `[🤮]`;
+        console.error(tag, whatWasThrown);
+        super(spaceTrim(`
+                Non-Error object was thrown
+
+                Note: Look for ${tag} in the console for more details
+                Please report issue on ${ADMIN_EMAIL}
+            `));
+        this.name = 'WrappedError';
+        Object.setPrototypeOf(this, WrappedError.prototype);
+    }
+}
+
+/**
+ * Helper used in catch blocks to assert that the error is an instance of `Error`
+ *
+ * @param whatWasThrown Any object that was thrown
+ * @returns Nothing if the error is an instance of `Error`
+ * @throws `WrappedError` or `UnexpectedError` if the error is not standard
+ *
+ * @private within the repository
+ */
+function assertsError(whatWasThrown) {
+    // Case 1: Handle error which was rethrown as `WrappedError`
+    if (whatWasThrown instanceof WrappedError) {
+        const wrappedError = whatWasThrown;
+        throw wrappedError;
+    }
+    // Case 2: Handle unexpected errors
+    if (whatWasThrown instanceof UnexpectedError) {
+        const unexpectedError = whatWasThrown;
+        throw unexpectedError;
+    }
+    // Case 3: Handle standard errors - keep them up to consumer
+    if (whatWasThrown instanceof Error) {
+        return;
+    }
+    // Case 4: Handle non-standard errors - wrap them into `WrappedError` and throw
+    throw new WrappedError(whatWasThrown);
+}
+
 /**
  * Checks if the value is [🚉] serializable as JSON
  * If not, throws an UnexpectedError with a rich error message and tracking
@@ -465,9 +515,7 @@ function checkSerializableAsJson(options) {
                 JSON.stringify(value); // <- TODO: [0]
             }
             catch (error) {
-                if (!(error instanceof Error)) {
-                    throw error;
-                }
+                assertsError(error);
                 throw new UnexpectedError(spaceTrim$1((block) => `
                             \`${name}\` is not serializable
 
@@ -689,13 +737,19 @@ function valueToString(value) {
             return value.toISOString();
         }
         else {
-            return JSON.stringify(value);
+            try {
+                return JSON.stringify(value);
+            }
+            catch (error) {
+                if (error instanceof TypeError && error.message.includes('circular structure')) {
+                    return VALUE_STRINGS.circular;
+                }
+                throw error;
+            }
         }
     }
     catch (error) {
-        if (!(error instanceof Error)) {
-            throw error;
-        }
+        assertsError(error);
         console.error(error);
         return VALUE_STRINGS.unserializable;
     }
@@ -1219,9 +1273,10 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'COMPLETION',
             modelTitle: 'davinci-002',
             modelName: 'davinci-002',
+            modelDescription: 'Legacy completion model with strong performance on text generation tasks. Optimized for complex instructions and longer outputs.',
             pricing: {
                 prompt: computeUsage(`$2.00 / 1M tokens`),
-                output: computeUsage(`$2.00 / 1M tokens`), // <- not sure
+                output: computeUsage(`$2.00 / 1M tokens`),
             },
         },
         /**/
@@ -1236,6 +1291,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo-16k',
             modelName: 'gpt-3.5-turbo-16k',
+            modelDescription: 'GPT-3.5 Turbo with extended 16k token context length for handling longer conversations and documents.',
             pricing: {
                 prompt: computeUsage(`$3.00 / 1M tokens`),
                 output: computeUsage(`$4.00 / 1M tokens`),
@@ -1259,6 +1315,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4',
             modelName: 'gpt-4',
+            modelDescription: 'GPT-4 is a powerful language model with enhanced reasoning, instruction-following capabilities, and 8K context window. Optimized for complex tasks requiring deep understanding.',
             pricing: {
                 prompt: computeUsage(`$30.00 / 1M tokens`),
                 output: computeUsage(`$60.00 / 1M tokens`),
@@ -1270,6 +1327,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4-32k',
             modelName: 'gpt-4-32k',
+            modelDescription: 'Extended context version of GPT-4 with a 32K token window for processing very long inputs and generating comprehensive responses for complex tasks.',
             pricing: {
                 prompt: computeUsage(`$60.00 / 1M tokens`),
                 output: computeUsage(`$120.00 / 1M tokens`),
@@ -1292,6 +1350,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4-turbo-2024-04-09',
             modelName: 'gpt-4-turbo-2024-04-09',
+            modelDescription: 'Latest stable GPT-4 Turbo model from April 2024 with enhanced reasoning and context handling capabilities. Offers 128K context window and improved performance.',
             pricing: {
                 prompt: computeUsage(`$10.00 / 1M tokens`),
                 output: computeUsage(`$30.00 / 1M tokens`),
@@ -1303,6 +1362,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo-1106',
             modelName: 'gpt-3.5-turbo-1106',
+            modelDescription: 'November 2023 version of GPT-3.5 Turbo with improved instruction following and a 16K token context window.',
             pricing: {
                 prompt: computeUsage(`$1.00 / 1M tokens`),
                 output: computeUsage(`$2.00 / 1M tokens`),
@@ -1314,6 +1374,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4-turbo',
             modelName: 'gpt-4-turbo',
+            modelDescription: 'More capable model than GPT-4 with improved instruction following, function calling and a 128K token context window for handling very large documents.',
             pricing: {
                 prompt: computeUsage(`$10.00 / 1M tokens`),
                 output: computeUsage(`$30.00 / 1M tokens`),
@@ -1325,6 +1386,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'COMPLETION',
             modelTitle: 'gpt-3.5-turbo-instruct-0914',
             modelName: 'gpt-3.5-turbo-instruct-0914',
+            modelDescription: 'September 2023 version of GPT-3.5 Turbo optimized for completion-style instruction following with a 4K context window.',
             pricing: {
                 prompt: computeUsage(`$1.50  / 1M tokens`),
                 output: computeUsage(`$2.00 / 1M tokens`), // <- For gpt-3.5-turbo-instruct
@@ -1336,6 +1398,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'COMPLETION',
             modelTitle: 'gpt-3.5-turbo-instruct',
             modelName: 'gpt-3.5-turbo-instruct',
+            modelDescription: 'Optimized version of GPT-3.5 for completion-style API with good instruction following and a 4K token context window.',
             pricing: {
                 prompt: computeUsage(`$1.50  / 1M tokens`),
                 output: computeUsage(`$2.00 / 1M tokens`),
@@ -1353,9 +1416,10 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo',
             modelName: 'gpt-3.5-turbo',
+            modelDescription: 'Latest version of GPT-3.5 Turbo with improved performance and instruction following capabilities. Default 4K context window with options for 16K.',
             pricing: {
-                prompt: computeUsage(`$3.00 / 1M tokens`),
-                output: computeUsage(`$6.00 / 1M tokens`), // <- Not sure, refer to gpt-3.5-turbo in Fine-tuning models
+                prompt: computeUsage(`$0.50 / 1M tokens`),
+                output: computeUsage(`$1.50 / 1M tokens`),
             },
         },
         /**/
@@ -1364,6 +1428,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo-0301',
             modelName: 'gpt-3.5-turbo-0301',
+            modelDescription: 'March 2023 version of GPT-3.5 Turbo with a 4K token context window. Legacy model maintained for backward compatibility.',
             pricing: {
                 prompt: computeUsage(`$1.50 / 1M tokens`),
                 output: computeUsage(`$2.00 / 1M tokens`),
@@ -1375,9 +1440,10 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'COMPLETION',
             modelTitle: 'babbage-002',
             modelName: 'babbage-002',
+            modelDescription: 'Efficient legacy completion model with a good balance of performance and speed. Suitable for straightforward text generation tasks.',
             pricing: {
                 prompt: computeUsage(`$0.40 / 1M tokens`),
-                output: computeUsage(`$0.40 / 1M tokens`), // <- Not sure
+                output: computeUsage(`$0.40 / 1M tokens`),
             },
         },
         /**/
@@ -1386,6 +1452,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4-1106-preview',
             modelName: 'gpt-4-1106-preview',
+            modelDescription: 'November 2023 preview version of GPT-4 Turbo with improved instruction following and a 128K token context window.',
             pricing: {
                 prompt: computeUsage(`$10.00 / 1M tokens`),
                 output: computeUsage(`$30.00 / 1M tokens`),
@@ -1397,6 +1464,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4-0125-preview',
             modelName: 'gpt-4-0125-preview',
+            modelDescription: 'January 2024 preview version of GPT-4 Turbo with improved reasoning capabilities and a 128K token context window.',
             pricing: {
                 prompt: computeUsage(`$10.00 / 1M tokens`),
                 output: computeUsage(`$30.00 / 1M tokens`),
@@ -1414,6 +1482,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo-0125',
             modelName: 'gpt-3.5-turbo-0125',
+            modelDescription: 'January 2024 version of GPT-3.5 Turbo with improved reasoning capabilities and a 16K token context window.',
             pricing: {
                 prompt: computeUsage(`$0.50 / 1M tokens`),
                 output: computeUsage(`$1.50  / 1M tokens`),
@@ -1425,9 +1494,10 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4-turbo-preview',
             modelName: 'gpt-4-turbo-preview',
+            modelDescription: 'Preview version of GPT-4 Turbo that points to the latest model version. Features improved instruction following, 128K token context window and lower latency.',
             pricing: {
                 prompt: computeUsage(`$10.00 / 1M tokens`),
-                output: computeUsage(`$30.00 / 1M tokens`), // <- Not sure, just for gpt-4-turbo
+                output: computeUsage(`$30.00 / 1M tokens`),
             },
         },
         /**/
@@ -1436,6 +1506,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'EMBEDDING',
             modelTitle: 'text-embedding-3-large',
             modelName: 'text-embedding-3-large',
+            modelDescription: "OpenAI's most capable text embedding model designed for high-quality embeddings for complex similarity tasks and information retrieval.",
             pricing: {
                 prompt: computeUsage(`$0.13  / 1M tokens`),
                 // TODO: [🏏] Leverage the batch API @see https://platform.openai.com/docs/guides/batch
@@ -1448,6 +1519,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'EMBEDDING',
             modelTitle: 'text-embedding-3-small',
             modelName: 'text-embedding-3-small',
+            modelDescription: 'Cost-effective embedding model with good performance for simpler tasks like text similarity and retrieval. Good balance of quality and efficiency.',
             pricing: {
                 prompt: computeUsage(`$0.02 / 1M tokens`),
                 // TODO: [🏏] Leverage the batch API @see https://platform.openai.com/docs/guides/batch
@@ -1460,6 +1532,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo-0613',
             modelName: 'gpt-3.5-turbo-0613',
+            modelDescription: 'June 2023 version of GPT-3.5 Turbo with function calling capabilities and a 4K token context window.',
             pricing: {
                 prompt: computeUsage(`$1.50 / 1M tokens`),
                 output: computeUsage(`$2.00 / 1M tokens`),
@@ -1471,6 +1544,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'EMBEDDING',
             modelTitle: 'text-embedding-ada-002',
             modelName: 'text-embedding-ada-002',
+            modelDescription: 'Legacy text embedding model suitable for text similarity and retrieval augmented generation use cases. Replaced by newer embedding-3 models.',
             pricing: {
                 prompt: computeUsage(`$0.1 / 1M tokens`),
                 // TODO: [🏏] Leverage the batch API @see https://platform.openai.com/docs/guides/batch
@@ -1501,11 +1575,11 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4o-2024-05-13',
             modelName: 'gpt-4o-2024-05-13',
+            modelDescription: 'May 2024 version of GPT-4o with enhanced multimodal capabilities, improved reasoning, and optimized for vision, audio and chat at lower latencies.',
             pricing: {
                 prompt: computeUsage(`$5.00 / 1M tokens`),
                 output: computeUsage(`$15.00 / 1M tokens`),
             },
-            //TODO: [main] !!3 Add gpt-4o-mini-2024-07-18 and all others to be up to date
         },
         /**/
         /**/
@@ -1513,6 +1587,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-4o',
             modelName: 'gpt-4o',
+            modelDescription: "OpenAI's most advanced multimodal model optimized for performance, speed, and cost. Capable of vision, reasoning, and high quality text generation.",
             pricing: {
                 prompt: computeUsage(`$5.00 / 1M tokens`),
                 output: computeUsage(`$15.00 / 1M tokens`),
@@ -1520,10 +1595,23 @@ const OPENAI_MODELS = exportJson({
         },
         /**/
         /**/
+        {
+            modelVariant: 'CHAT',
+            modelTitle: 'gpt-4o-mini',
+            modelName: 'gpt-4o-mini',
+            modelDescription: 'Smaller, more cost-effective version of GPT-4o with good performance across text, vision, and audio tasks at reduced complexity.',
+            pricing: {
+                prompt: computeUsage(`$3.00 / 1M tokens`),
+                output: computeUsage(`$9.00 / 1M tokens`),
+            },
+        },
+        /**/
+        /**/
         {
             modelVariant: 'CHAT',
             modelTitle: 'o1-preview',
             modelName: 'o1-preview',
+            modelDescription: 'Advanced reasoning model with exceptional performance on complex logical, mathematical, and analytical tasks. Built for deep reasoning and specialized professional tasks.',
             pricing: {
                 prompt: computeUsage(`$15.00 / 1M tokens`),
                 output: computeUsage(`$60.00 / 1M tokens`),
@@ -1535,6 +1623,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'o1-preview-2024-09-12',
             modelName: 'o1-preview-2024-09-12',
+            modelDescription: 'September 2024 version of O1 preview with specialized reasoning capabilities for complex tasks requiring precise analytical thinking.',
             //             <- TODO: [💩] Some better system to organize theese date suffixes and versions
             pricing: {
                 prompt: computeUsage(`$15.00 / 1M tokens`),
@@ -1547,6 +1636,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'o1-mini',
             modelName: 'o1-mini',
+            modelDescription: 'Smaller, cost-effective version of the O1 model with good performance on reasoning tasks while maintaining efficiency for everyday analytical use.',
             pricing: {
                 prompt: computeUsage(`$3.00 / 1M tokens`),
                 output: computeUsage(`$12.00 / 1M tokens`),
@@ -1558,10 +1648,10 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'o1',
             modelName: 'o1',
+            modelDescription: "OpenAI's advanced reasoning model focused on logic and problem-solving. Designed for complex analytical tasks with rigorous step-by-step reasoning. 128K context window.",
             pricing: {
-                prompt: computeUsage(`$3.00 / 1M tokens`),
-                output: computeUsage(`$12.00 / 1M tokens`),
-                // <- TODO: !! Unsure, check the pricing
+                prompt: computeUsage(`$15.00 / 1M tokens`),
+                output: computeUsage(`$60.00 / 1M tokens`),
             },
         },
         /**/
@@ -1570,6 +1660,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'o3-mini',
             modelName: 'o3-mini',
+            modelDescription: 'Cost-effective reasoning model optimized for academic and scientific problem-solving. Efficient performance on STEM tasks with deep mathematical and scientific knowledge. 128K context window.',
             pricing: {
                 prompt: computeUsage(`$3.00 / 1M tokens`),
                 output: computeUsage(`$12.00 / 1M tokens`),
@@ -1582,6 +1673,7 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'o1-mini-2024-09-12',
             modelName: 'o1-mini-2024-09-12',
+            modelDescription: "September 2024 version of O1-mini with balanced reasoning capabilities and cost-efficiency. Good for analytical tasks that don't require the full O1 model.",
             pricing: {
                 prompt: computeUsage(`$3.00 / 1M tokens`),
                 output: computeUsage(`$12.00 / 1M tokens`),
@@ -1593,12 +1685,14 @@ const OPENAI_MODELS = exportJson({
             modelVariant: 'CHAT',
             modelTitle: 'gpt-3.5-turbo-16k-0613',
             modelName: 'gpt-3.5-turbo-16k-0613',
+            modelDescription: 'June 2023 version of GPT-3.5 Turbo with extended 16k token context window for processing longer conversations and documents.',
             pricing: {
                 prompt: computeUsage(`$3.00 / 1M tokens`),
                 output: computeUsage(`$4.00 / 1M tokens`),
             },
         },
         /**/
+        // <- [🕕]
     ],
 });
 /**
@@ -1660,6 +1754,9 @@ resultContent, rawResponse) {
  * TODO: [🤝] DRY Maybe some common abstraction between `computeOpenAiUsage` and `computeAnthropicClaudeUsage`
  */
 
+// Default rate limits (requests per minute) - adjust as needed based on OpenAI tier
+const DEFAULT_RPM = 60;
+// <- TODO: !!! Put in some better place
 /**
  * Execution Tools for calling OpenAI API
  *
@@ -1677,6 +1774,10 @@ class OpenAiExecutionTools {
          * OpenAI API client.
          */
         this.client = null;
+        // TODO: Allow configuring rate limits via options
+        this.limiter = new Bottleneck({
+            minTime: 60000 / (this.options.maxRequestsPerMinute || DEFAULT_RPM),
+        });
     }
     get title() {
         return 'OpenAI';
@@ -1780,7 +1881,10 @@ class OpenAiExecutionTools {
         if (this.options.isVerbose) {
             console.info(colors.bgWhite('rawRequest'), JSON.stringify(rawRequest, null, 4));
         }
-        const rawResponse = await client.chat.completions.create(rawRequest).catch((error) => {
+        const rawResponse = await this.limiter
+            .schedule(() => client.chat.completions.create(rawRequest))
+            .catch((error) => {
+            assertsError(error);
             if (this.options.isVerbose) {
                 console.info(colors.bgRed('error'), error);
             }
@@ -1856,7 +1960,10 @@ class OpenAiExecutionTools {
         if (this.options.isVerbose) {
             console.info(colors.bgWhite('rawRequest'), JSON.stringify(rawRequest, null, 4));
         }
-        const rawResponse = await client.completions.create(rawRequest).catch((error) => {
+        const rawResponse = await this.limiter
+            .schedule(() => client.completions.create(rawRequest))
+            .catch((error) => {
+            assertsError(error);
             if (this.options.isVerbose) {
                 console.info(colors.bgRed('error'), error);
             }
@@ -1919,7 +2026,10 @@ class OpenAiExecutionTools {
         if (this.options.isVerbose) {
             console.info(colors.bgWhite('rawRequest'), JSON.stringify(rawRequest, null, 4));
         }
-        const rawResponse = await client.embeddings.create(rawRequest).catch((error) => {
+        const rawResponse = await this.limiter
+            .schedule(() => client.embeddings.create(rawRequest))
+            .catch((error) => {
+            assertsError(error);
             if (this.options.isVerbose) {
                 console.info(colors.bgRed('error'), error);
             }