Modification of AWS Bedrock to Amazon Bedrock

Author: ronyblum

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -24,7 +24,7 @@ body:
         - OpenAI
         - OpenAI Compatible
         - GCP Vertex AI
-        - AWS Bedrock
+        - Amazon Bedrock
         - Requesty
         - Glama
         - VS Code LM API

cline_docs/bedrock/bedrock-cache-strategy-documentation.md

@@ -1,6 +1,6 @@
 # Cache Strategy Documentation
 
-This document provides an overview of the cache strategy implementation for AWS Bedrock in the Roo-Code project, including class relationships and sequence diagrams.
+This document provides an overview of the cache strategy implementation for Amazon Bedrock in the Roo-Code project, including class relationships and sequence diagrams.
 
 ## Class Relationship Diagram
 
@@ -89,7 +89,7 @@ sequenceDiagram
     participant Client as Client Code
     participant Bedrock as AwsBedrockHandler
     participant Strategy as MultiPointStrategy
-    participant AWS as AWS Bedrock Service
+    participant AWS as Amazon Bedrock Service
 
     Client->>Bedrock: createMessage(systemPrompt, messages)
     Note over Bedrock: Generate conversationId to track cache points
@@ -143,7 +143,7 @@ sequenceDiagram
 
 ### Cache Strategy
 
-The cache strategy system is designed to optimize the placement of cache points in AWS Bedrock API requests. Cache points allow the service to reuse previously processed parts of the prompt, reducing token usage and improving response times.
+The cache strategy system is designed to optimize the placement of cache points in Amazon Bedrock API requests. Cache points allow the service to reuse previously processed parts of the prompt, reducing token usage and improving response times.
 
 - **MultiPointStrategy**: Upon first MR of Bedrock caching, this strategy is used for all cache point placement scenarios. It distributes cache points throughout the conversation to maximize caching efficiency, whether the model supports one or multiple cache points.
 
@@ -180,14 +180,14 @@ The simplified approach ensures that:
 
 The examples in this document reflect this optimized implementation.
 
-### Integration with AWS Bedrock
+### Integration with Amazon Bedrock
 
 The AwsBedrockHandler class integrates with the cache strategies by:
 
 1. Determining if the model supports prompt caching
 2. Creating the appropriate strategy based on model capabilities
 3. Applying the strategy to format messages with optimal cache points
-4. Sending the formatted request to AWS Bedrock
+4. Sending the formatted request to Amazon Bedrock
 5. Processing and returning the response
 
 ## Usage Considerations

cline_docs/bedrock/model-identification.md

@@ -1,6 +1,6 @@
 # Bedrock Model Identification
 
-This document explains how model information is identified and managed in the AWS Bedrock provider implementation (`bedrock.ts`). It focuses on the sequence of operations that determine the `costModelConfig` property, which is crucial for token counting, pricing, and other features.
+This document explains how model information is identified and managed in the Amazon Bedrock provider implementation (`bedrock.ts`). It focuses on the sequence of operations that determine the `costModelConfig` property, which is crucial for token counting, pricing, and other features.
 
 ## Model Identification Flow
 

evals/packages/types/src/roo-code.ts

@@ -312,7 +312,7 @@ export const providerSettingsSchema = z.object({
 	openRouterBaseUrl: z.string().optional(),
 	openRouterSpecificProvider: z.string().optional(),
 	openRouterUseMiddleOutTransform: z.boolean().optional(),
-	// AWS Bedrock
+	// Amazon Bedrock
 	awsAccessKey: z.string().optional(),
 	awsSecretKey: z.string().optional(),
 	awsSessionToken: z.string().optional(),
@@ -403,7 +403,7 @@ const providerSettingsRecord: ProviderSettingsRecord = {
 	openRouterBaseUrl: undefined,
 	openRouterSpecificProvider: undefined,
 	openRouterUseMiddleOutTransform: undefined,
-	// AWS Bedrock
+	// Amazon Bedrock
 	awsAccessKey: undefined,
 	awsSecretKey: undefined,
 	awsSessionToken: undefined,

src/api/providers/bedrock.ts

@@ -22,7 +22,7 @@ import { Message, SystemContentBlock } from "@aws-sdk/client-bedrock-runtime"
 // New cache-related imports
 import { MultiPointStrategy } from "../transform/cache-strategy/multi-point-strategy"
 import { ModelInfo as CacheModelInfo } from "../transform/cache-strategy/types"
-import { AWS_BEDROCK_REGION_INFO } from "../../shared/aws_regions"
+import { AMAZON_BEDROCK_REGION_INFO } from "../../shared/aws_regions"
 
 const BEDROCK_DEFAULT_TEMPERATURE = 0.3
 const BEDROCK_MAX_TOKENS = 4096
@@ -495,7 +495,7 @@ export class AwsBedrockHandler extends BaseProvider implements SingleCompletionH
 
 	private parseArn(arn: string, region?: string) {
 		/*
-		 * VIA Roo analysis: platform-independent Regex. It's designed to parse AWS Bedrock ARNs and doesn't rely on any platform-specific features
+		 * VIA Roo analysis: platform-independent Regex. It's designed to parse Amazon Bedrock ARNs and doesn't rely on any platform-specific features
 		 * like file path separators, line endings, or case sensitivity behaviors. The forward slashes in the regex are properly escaped and
 		 * represent literal characters in the AWS ARN format, not filesystem paths. This regex will function consistently across Windows,
 		 * macOS, Linux, and any other operating system where JavaScript runs.
@@ -562,7 +562,7 @@ export class AwsBedrockHandler extends BaseProvider implements SingleCompletionH
 			region: undefined,
 			modelType: undefined,
 			modelId: undefined,
-			errorMessage: "Invalid ARN format. ARN should follow the AWS Bedrock ARN pattern.",
+			errorMessage: "Invalid ARN format. ARN should follow the Amazon Bedrock ARN pattern.",
 			crossRegionInference: false,
 		}
 	}
@@ -700,16 +700,16 @@ export class AwsBedrockHandler extends BaseProvider implements SingleCompletionH
 
 	/************************************************************************************
 	 *
-	 *     AWS REGIONS
+	 *     AMAZON REGIONS
 	 *
 	 *************************************************************************************/
 
 	private static getPrefixList(): string[] {
-		return Object.keys(AWS_BEDROCK_REGION_INFO)
+		return Object.keys(AMAZON_BEDROCK_REGION_INFO)
 	}
 
 	private static getPrefixForRegion(region: string): string | undefined {
-		for (const [prefix, info] of Object.entries(AWS_BEDROCK_REGION_INFO)) {
+		for (const [prefix, info] of Object.entries(AMAZON_BEDROCK_REGION_INFO)) {
 			if (info.pattern && region.startsWith(info.pattern)) {
 				return prefix
 			}
@@ -718,7 +718,7 @@ export class AwsBedrockHandler extends BaseProvider implements SingleCompletionH
 	}
 
 	private static prefixIsMultiRegion(arnPrefix: string): boolean {
-		for (const [prefix, info] of Object.entries(AWS_BEDROCK_REGION_INFO)) {
+		for (const [prefix, info] of Object.entries(AMAZON_BEDROCK_REGION_INFO)) {
 			if (arnPrefix === prefix) {
 				if (info?.multiRegion) return info.multiRegion
 				else return false
@@ -791,7 +791,7 @@ Suggestions:
 2. Split your request into smaller chunks
 3. Use a model with a larger context window
 4. If rate limited, reduce request frequency
-5. Check your AWS Bedrock quotas and limits`,
+5. Check your Amazon Bedrock quotas and limits`,
 			logLevel: "error",
 		},
 		ON_DEMAND_NOT_SUPPORTED: {

src/schemas/index.ts

@@ -319,7 +319,7 @@ export const providerSettingsSchema = z.object({
 	openRouterBaseUrl: z.string().optional(),
 	openRouterSpecificProvider: z.string().optional(),
 	openRouterUseMiddleOutTransform: z.boolean().optional(),
-	// AWS Bedrock
+	// Amazon Bedrock
 	awsAccessKey: z.string().optional(),
 	awsSecretKey: z.string().optional(),
 	awsSessionToken: z.string().optional(),
@@ -414,7 +414,7 @@ const providerSettingsRecord: ProviderSettingsRecord = {
 	openRouterBaseUrl: undefined,
 	openRouterSpecificProvider: undefined,
 	openRouterUseMiddleOutTransform: undefined,
-	// AWS Bedrock
+	// Amazon Bedrock
 	awsAccessKey: undefined,
 	awsSecretKey: undefined,
 	awsSessionToken: undefined,

src/shared/api.ts

@@ -77,8 +77,8 @@ export const anthropicModels = {
 		cacheReadsPrice: 0.03,
 	},
 } as const satisfies Record<string, ModelInfo> // as const assertion makes the object deeply readonly
-
-// AWS Bedrock
+// Amazon Bedrock
+// https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html
 // https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html
 export interface MessageContent {
 	type: "text" | "image" | "video" | "tool_use" | "tool_result"

src/shared/aws_regions.ts

@@ -2,7 +2,7 @@
  * AWS Region information mapping
  * Maps region prefixes to their full region IDs and descriptions
  */
-export const AWS_BEDROCK_REGION_INFO: Record<
+export const AMAZON_BEDROCK_REGION_INFO: Record<
 	string,
 	{
 		regionId: string
@@ -69,7 +69,7 @@ export const AWS_BEDROCK_REGION_INFO: Record<
 }
 
 // Extract unique region IDs from REGION_INFO and create the AWS_REGIONS array
-export const AWS_REGIONS = Object.values(AWS_BEDROCK_REGION_INFO)
+export const AWS_REGIONS = Object.values(AMAZON_BEDROCK_REGION_INFO)
 	// Extract all region IDs
 	.map((info) => ({ value: info.regionId, label: info.regionId }))
 	// Filter to unique region IDs (remove duplicates)

webview-ui/src/components/settings/constants.ts

@@ -28,7 +28,7 @@ export const PROVIDERS = [
 	{ value: "openai-native", label: "OpenAI" },
 	{ value: "openai", label: "OpenAI Compatible" },
 	{ value: "vertex", label: "GCP Vertex AI" },
-	{ value: "bedrock", label: "AWS Bedrock" },
+	{ value: "bedrock", label: "Amazon Bedrock" },
 	{ value: "glama", label: "Glama" },
 	{ value: "vscode-lm", label: "VS Code LM API" },
 	{ value: "mistral", label: "Mistral" },

webview-ui/src/i18n/locales/ca/settings.json

@@ -101,7 +101,7 @@
 		"searchPlaceholder": "Cerca perfils",
 		"noMatchFound": "No s'han trobat perfils coincidents",
 		"vscodeLmDescription": "L'API del model de llenguatge de VS Code us permet executar models proporcionats per altres extensions de VS Code (incloent-hi, però no limitat a, GitHub Copilot). La manera més senzilla de començar és instal·lar les extensions Copilot i Copilot Chat des del VS Code Marketplace.",
-		"awsCustomArnUse": "Introduïu un ARN vàlid d'AWS Bedrock per al model que voleu utilitzar. Exemples de format:",
+		"awsCustomArnUse": "Introduïu un ARN vàlid d'Amazon Bedrock per al model que voleu utilitzar. Exemples de format:",
 		"awsCustomArnDesc": "Assegureu-vos que la regió a l'ARN coincideix amb la regió d'AWS seleccionada anteriorment.",
 		"apiKeyStorageNotice": "Les claus API s'emmagatzemen de forma segura a l'Emmagatzematge Secret de VSCode",
 		"useCustomBaseUrl": "Utilitzar URL base personalitzada",
@@ -403,7 +403,7 @@
 	},
 	"validation": {
 		"apiKey": "Heu de proporcionar una clau API vàlida.",
-		"awsRegion": "Heu de triar una regió per utilitzar AWS Bedrock.",
+		"awsRegion": "Heu de triar una regió per utilitzar Amazon Bedrock.",
 		"googleCloud": "Heu de proporcionar un ID de projecte i regió de Google Cloud vàlids.",
 		"modelId": "Heu de proporcionar un ID de model vàlid.",
 		"modelSelector": "Heu de proporcionar un selector de model vàlid.",