Bedrock L2 Construct #688

dineshSajwan · 2025-01-27T18:35:09Z

This is a request for Amazon Bedrock L2 construct that simplifies the creation of multiple Bedrock features by providing a wrapper over the Bedrock L1 construct. It exposes functions that enable users to create features with minimal code. Key features include Bedrock Agent, Knowledge Base, Guardrails, Inference Profiles, and Prompt. See #686 for
additional details.

APIs are signed off by @{BAR_RAISER}.

By submitting this pull request, I confirm that my contribution is made under
the terms of the Apache-2.0 license

hrudu-dev

Overall, the document is well-structured and provides a comprehensive overview of the Bedrock L2 construct. However, enhancing clarity in certain sections and ensuring completeness will greatly benefit users.

hrudu-dev · 2025-02-04T20:04:30Z

text/0686-bedrock-l2-construct.md

+
+The Bedrock L2 construct simplifies the creation of multiple Bedrock features by providing a wrapper over the Bedrock L1 construct. It exposes functions that enable users to create features with minimal code. Key features include Bedrock Agent, Knowledge Base, Guardrails, Inference Profiles, and Prompt.
+
+


Please ensure that all sections of the new construct are complete and provide sufficient details for users to understand and implement the construct.

thanks @hrudu-dev , yes we have added all the construct props , API documentation and initializer code for all the resources.

hrudu-dev · 2025-02-04T20:04:46Z

text/0686-bedrock-l2-construct.md

+For more details please refer here [Amazon Bedrock README](https://github.com/awslabs/generative-ai-cdk-constructs/blob/main/src/cdk-lib/bedrock/README.md).
+
+
+## Knowledge Base


Verify that the documentation (README) is clear and includes examples for each feature of the Bedrock L2 construct. It should also explain how to integrate with other AWS services.Verify that the documentation (README) is clear and includes examples for each feature of the Bedrock L2 construct. It should also explain how to integrate with other AWS services.

Yes, the readme has information on how to create , KB with different vector stores and data sources with different strategies. same for agent , prompt and inference profiles. for details example we have another repo under aws-samples org which use these construct and create a full fledge solution.

hrudu-dev · 2025-02-04T20:05:08Z

text/0686-bedrock-l2-construct.md

+const auroraDb = aurora.AmazonAuroraVectorStore.fromExistingAuroraVectorStore(stack, 'ExistingAuroraVectorStore', {
+  clusterIdentifier: 'aurora-serverless-vector-cluster',
+  databaseName: 'bedrock_vector_db',
+  schemaName: 'bedrock_integration',


Confirm that all chunking strategies are thoroughly explained and example code is provided for each (Fixed Size, Hierarchical, Semantic, etc.

hrudu-dev · 2025-02-04T20:05:28Z

text/0686-bedrock-l2-construct.md

+
+const kb = new KnowledgeBase(stack, 'MyKnowledgeBase', {
+  name: 'MyKnowledgeBase',
+  embeddingsModel: BedrockFoundationModel.COHERE_EMBED_MULTILINGUAL_V3,


Review the guardrails to ensure that they cover all necessary areas for content filtering, denied topics, word filtering, and sensitive information filters.

hrudu-dev · 2025-02-04T20:05:42Z

text/0686-bedrock-l2-construct.md

+
+- **Fixed Size Chunking**: This method divides the data into fixed-size chunks, with each chunk
+  containing a predetermined number of tokens. This strategy is useful when the data is uniform
+  in size and structure.


Suggest adhering to AWS best practices, especially in areas involving security (e.g., handling AWS Secrets Manager, IAM roles).

hrudu-dev · 2025-02-04T20:06:16Z

text/0686-bedrock-l2-construct.md

+
+const variant2 = PromptVariant.text({
+  variantName: "variant2",
+  model: claudeModel,


Consider adding a section on how users can provide feedback or report issues with the Bedrock L2 construct.

once the construct is pushed to aws-cdk repo, one can open an issue for the construct in that repo.

hrudu-dev · 2025-02-04T20:06:27Z

text/0686-bedrock-l2-construct.md

+
+## Prompt management
+
+Amazon Bedrock provides the ability to create and save prompts using Prompt management so that you can save


Clarify the versioning strategy for the constructs. It’s important for users to know how to manage updates and changes.

Example of PromptVariant describe the same.

hrudu-dev · 2025-02-04T20:06:41Z

text/0686-bedrock-l2-construct.md

+  blockedOutputsMessaging: 'blockedOutputsMessaging',
+  name: 'namemycfnguardrails',
+  wordPolicyConfig: {
+    wordsConfig: [


Please include a section on testing the Bedrock L2 construct to ensure users understand how to validate their implementations.

The code snippet added shows how to use the construct. The final code will have unit test and integ test to validate the implementation.

hrudu-dev · 2025-02-04T20:06:53Z

text/0686-bedrock-l2-construct.md

+
+#### Specific version
+
+You can use the `AgentAlias` resource if you want to create an Alias for an existing Agent.


It would be beneficial to include error handling examples in the documentation to guide users on how to manage potential issues.

The construct error handling return appropriate user message at the time of cdk synth.

hrudu-dev · 2025-02-04T20:07:05Z

text/0686-bedrock-l2-construct.md

+### Create an Agent
+
+The following example creates an Agent with a simple instruction and default prompts that consults a Knowledge Base.
+


Ensure that all example code snippets are well-commented and easy to follow. Consider adding more context or explanations for complex sections.

noted. with code checkin we can review that again.

hrudu-dev · 2025-02-04T20:09:36Z

I've reviewed the documentation and it looks great! Happy to contribute further if needed. Thank you!

xazhao · 2025-02-05T18:21:11Z

text/0686-bedrock-l2-construct.md

+
+### What are the drawbacks of this solution?
+
+The Knowledge Base vector stores (OpenSearch and Aurora clusters) utilize custom resource lambda functions, as there are no underlying L1 constructs available.


I see there's CFN resource for OpenSearch collection with vector search. https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-resource-opensearchserverless-collection.html

Is it different that what need here?

Yes, we have CFN resource for vector collection but not for vector index creation. The custom resource used here create a vector index. But like we discussed we can skip the vector index part from the construct and enable it to use an existing collection and index.

xazhao · 2025-02-05T18:21:39Z

text/0686-bedrock-l2-construct.md

+const importedGuardrail = bedrock.Guardrail.fromCfnGuardrail(cfnGuardrail);
+```
+
+Python


We can remove the Python code here in readme and RFC.

Done, will push with next commit.

xazhao · 2025-02-05T18:25:21Z

text/0686-bedrock-l2-construct.md

+
+### Why should we _not_ do this?
+
+The construct is published via the [generative-ai-cdk-constructs](https://github.com/awslabs/generative-ai-cdk-constructs/blob/main/src/cdk-lib/bedrock/README.md) repository. However, due to the increasing demands and expanding use cases, maintaining it within this repository has become challenging.


maintaining it within this repository has become challenging.

Why maintaining it within this repo is more challenging than merging into aws-cdk?

aws-gen-ai-cdk-construct was an experimental initiative with a small team. We feel aws-cdk is the right place for this construct on long term.

xazhao · 2025-02-05T22:07:00Z

text/0686-bedrock-l2-construct.md

+```
+
+
+#### Knowledge Base - Data Sources


Do these APIs just create AWS::Bedrock::DataSource CFN resources?

xazhao · 2025-02-12T20:29:45Z

text/0686-bedrock-l2-construct.md

+* **API Bar Raiser**: @{BAR_RAISER_USER}
+
+
+The Bedrock L2 construct simplifies the creation of multiple Bedrock features by providing a wrapper over the Bedrock L1 construct. It exposes functions that enable users to create features with minimal code. Key features include Bedrock Agent, Knowledge Base, Guardrails, Inference Profiles, and Prompt.


How does it simplify the creation of Bedrock features? Could you provide a comparison between using L1 and L2?

ack, updated the readme - https://github.com/dineshSajwan/aws-cdk-rfcs/blob/main/text/0686-bedrock-l2-construct.md#bedrock-l2-construct

xazhao · 2025-02-12T20:30:50Z

text/0686-bedrock-l2-construct.md

+* **API Bar Raiser**: @{BAR_RAISER_USER}
+
+
+The Bedrock L2 construct simplifies the creation of multiple Bedrock features by providing a wrapper over the Bedrock L1 construct. It exposes functions that enable users to create features with minimal code. Key features include Bedrock Agent, Knowledge Base, Guardrails, Inference Profiles, and Prompt.


Knowledge Base is not in the scope of this RFC and L2 at the moment

Apologies for any confusion but Knowledge base can still be the part of the bedrock L2 construct . But the vector store based KB will not be part of the scope since it needs a custom resource to create an index. For this I have updated the documentation create vector store knowledge . Basically, the construct will expect an existing vector store . It supports opensearch serverless, amazon aurora and pinecone as existing vector store. This approach will have no custom resource in bedrock L2 construct.

xazhao · 2025-02-12T20:37:14Z

text/0686-bedrock-l2-construct.md

+
+The resource accepts an `instruction` prop that is provided to any Bedrock Agent it is associated with so the agent can decide when to query the Knowledge Base.
+
+Amazon Bedrock Knowledge Bases currently only supports S3 as a data source. The `S3DataSource` resource is used to configure how the Knowledge Base handles the data source.


only supports S3 as a data source.
In the section Data Source, it mentions there are multiple data source supported. What's the difference?

Typo error. removed it in 93e54fd

xazhao · 2025-02-12T20:38:18Z

text/0686-bedrock-l2-construct.md

+
+```ts
+import * as s3 from 'aws-cdk-lib/aws-s3';
+import { bedrock } from '@cdklabs/generative-ai-cdk-constructs';


Mention why we need to use an external construct.

This is a typo. Replaced it with import { bedrock } from 'aws-cdk-lib/aws-bedrock';. Eventually when the construct will be pushed to aws-cdk-lib it can be imported in this fashion. 93e54fd

xazhao · 2025-02-12T20:40:05Z

text/0686-bedrock-l2-construct.md

+
+Amazon Bedrock Knowledge Bases enable you to provide foundation models and agents with contextual information from your company’s private data sources. This enhances the relevance, accuracy, and customization of their responses.
+
+### Create a Knowledge Base


Since this is currently out of scope for Bedrock L2 in CDK, we can make this section short and refer to the readme in the other repo

Apologies for any confusion but Knowledge base can still be the part of the bedrock L2 construct . But the vector store based KB will not be part of the scope since it needs a custom resource to create an index. For this I have updated the documentation create vector store knowledge . Basically, the construct will expect an existing vector store . It supports opensearch serverless, amazon aurora and pinecone as existing vector store. This approach will have no custom resource in bedrock L2 construct.

xazhao · 2025-02-12T20:50:54Z

text/0686-bedrock-l2-construct.md

+- **Amazon S3**. You can either create a new data source using the `bedrock.S3DataSource(..)` class, or using the
+  `kb.addS3DataSource(..)`.


Which is the recommended way? IMO kb.addS3DataSource() is a better API. If data source is not a stand alone resource, we can just keep the add method.

It depends. If someone want to use their existing kb and want to associate a datasource using construct, they can use bedrock.S3Datasource. If they have an existing datasource and want to use kb using construct they can use kb.addS3DataSource. we have seen demand for both scenarios.

xazhao · 2025-02-12T20:54:05Z

text/0686-bedrock-l2-construct.md

+Content filters – Adjust filter strengths to block input prompts or model responses containing harmful content.
+
+Denied topics – Define a set of topics that are undesirable in the context of your application. These topics will be blocked if detected in user queries or model responses.
+
+Word filters – Configure filters to block undesirable words, phrases, and profanity. Such words can include offensive terms, competitor names etc.
+
+Sensitive information filters – Block or mask sensitive information such as personally identifiable information (PII) or custom regex in user inputs and model responses.


Let's add some bullet points here so it's easier to read.

Done - 93e54fd

xazhao · 2025-02-12T21:05:07Z

text/0686-bedrock-l2-construct.md

+const app = new cdk.App();
+const stack = new cdk.Stack(app, 'aws-cdk-bedrock-data-sources-integ-test');


These should be removed. Also we should keep the code example as minimal as possible. Ideally it should only includes required property unless that property is for a specific feature.

Done - 01e9249

Pull request has been modified.

xazhao · 2025-02-14T23:17:55Z

text/0686-bedrock-l2-construct.md

+Amazon Bedrock Knowledge Bases provides foundation models and agents with contextual data from private sources. This capability enhances
+response relevance, accuracy, and customization for your specific use cases.
+
+## Supported Knowledge Base Types


nit: this should be ###

xazhao · 2025-02-14T23:20:08Z

text/0686-bedrock-l2-construct.md

+
+2. `instruction` prop: Provided to associated Bedrock Agents to determine when to query the Knowledge Base
+
+3. embeddingsModel: Foundation model supported with bedrock.


Suggested change

3. embeddingsModel: Foundation model supported with bedrock.

3. `embeddingsModel` prop: Foundation model supported with bedrock.

xazhao · 2025-02-14T23:23:43Z

text/0686-bedrock-l2-construct.md

+
+To create a vector knowledge Base, a vector index on a vector store is required. The resource accepts:
+
+1. `storageConfiguration` prop: An existing vector store from:


Looking at https://docs.aws.amazon.com/bedrock/latest/APIReference/API_agent_StorageConfiguration.html, there are 2 more configuration supported: mongoDbAtlasConfiguration and redisEnterpriseCloudConfiguration. Are they supported?

No , not yet, but may be in Future.

xazhao · 2025-02-14T23:25:20Z

text/0686-bedrock-l2-construct.md

+  embeddingsModel: bedrock.BedrockFoundationModel.TITAN_EMBED_TEXT_V1,
+  instruction: 'Use this knowledge base to answer questions about books. ' + 'It contains the full text of novels.',
+  storageConfiguration:[{
+    type:'OPENSEARCH_SERVERLESS',


Later in the implementation, it should be a static variable.

Agreed. Noted.

xazhao · 2025-02-14T23:26:33Z

text/0686-bedrock-l2-construct.md

+N
+ote - you need to provide clusterIdentifier, databaseName, vpc, secret and auroraSecurityGroupId used in


nit: extra line here.

xazhao · 2025-02-14T23:37:27Z

text/0686-bedrock-l2-construct.md

+const docBucket = new s3.Bucket(this, "DocBucket");
+
+new bedrock.S3DataSource(this, "DataSource", {
+  bucket: docBucket,
+  knowledgeBase: kb,
+  dataSourceName: "books",
+  chunkingStrategy: bedrock.ChunkingStrategy.FIXED_SIZE,
+});
+```


xazhao · 2025-02-14T23:38:09Z

text/0686-bedrock-l2-construct.md

+import * as s3 from 'aws-cdk-lib/aws-s3';
+import { pinecone, bedrock } from 'aws-cdk-lib/aws-bedrock';
+
+const pineconeds = new pinecone.PineconeVectorStore({


Removed. same as AmazonAuroraVectorStore it can go as separate construct and then add a from method in bedrock construct.

xazhao · 2025-02-14T23:38:30Z

text/0686-bedrock-l2-construct.md

+const docBucket = new s3.Bucket(this, 'DocBucket');
+
+new bedrock.S3DataSource(this, 'DataSource', {
+  bucket: docBucket,
+  knowledgeBase: kb,
+  dataSourceName: 'books',
+  chunkingStrategy: bedrock.ChunkingStrategy.FIXED_SIZE,
+});
+```


Same. We don't need data source code here.

xazhao · 2025-02-14T23:40:55Z

text/0686-bedrock-l2-construct.md

+kb.addWebCrawlerDataSource({
+  sourceUrls: ['https://docs.aws.amazon.com/'],
+  chunkingStrategy: ChunkingStrategy.HIERARCHICAL_COHERE,
+  customTransformation: CustomTransformation.lambda({
+    lambdaFunction: lambdaFunction,
+    s3BucketUri: `s3://${bucket.bucketName}/chunk-processor/`,
+  }),
+});


Instead of putting examples all together. Can we put them individual section above? It would be easier to read.

xazhao · 2025-02-14T23:47:00Z

text/0686-bedrock-l2-construct.md

+```ts
+CustomTransformation.lambda({
+lambdaFunction: lambdaFunction,
+s3BucketUri: `s3://${bucket.bucketName}/chunk-processor/`,
+}),
+```


nit: format

xazhao · 2025-02-18T23:16:33Z

text/0686-bedrock-l2-construct.md

+
+### Prepare the Agent
+
+he Agent constructs take an optional parameter shouldPrepareAgent to indicate that the Agent should be prepared after any updates to


xazhao · 2025-02-18T23:32:11Z

text/0686-bedrock-l2-construct.md

+const agentAlias2 = new bedrock.AgentAlias(this, 'myalias2', {
+  aliasName: 'myalias',
+  agent: agent,
+  agentVersion: '1', // optional
+  description: 'mytest'
+});


This seems to be 1:1 mapping to L1 resource. IMO we can replace it with an API agent.addAlias(...). What do you think?

We started with addAlias method initially but then we faced some issue and then had to switch to this way of creation. So instead of agent adding alias the control is with alias to associate an agent to it. This class apart from initializing agent also support method like fromAttributes. This method can import and existing alias created outside of CDK. Added it in new commit.

What's the issue with addAlias method? AgentAlias is essentially the same as L1 and if we need it, we can just use L1. For the import, I think it's a very edge case where agent is created in the stack while importing an alias created somewhere else.

The fromAttributes is a must to handle the default test alias of bedrock agent (created by the service itself).
Can't recall the issues with addAlias method , I think we can bring it back and verify there are no issue because of it.

sounds good.

text/0686-bedrock-l2-construct.md

dinsajwa added 2 commits January 27, 2025 11:11

feat(bedorck): rfc for bedrockL2 construct

cc1f5bc

feat(bedorck): rfc for bedrockL2 construct

633baa7

SimonCMoore mentioned this pull request Feb 4, 2025

AWS Bedrock Integration Working Group #691

Open

hrudu-dev reviewed Feb 4, 2025

View reviewed changes

xazhao reviewed Feb 6, 2025

View reviewed changes

dinsajwa and others added 2 commits February 12, 2025 10:02

added review comments

2615062

Merge branch 'main' into main

9d295d4

xazhao previously requested changes Feb 12, 2025

View reviewed changes

xazhao reviewed Feb 12, 2025

View reviewed changes

dinsajwa added 2 commits February 12, 2025 18:34

added kendra kb , review comments and updated lint warnings

93e54fd

Merge branch 'main' of https://github.com/dineshSajwan/aws-cdk-rfcs

6f52fad

dinsajwa added 2 commits February 12, 2025 18:40

removed unwanted code

01e9249

added kb design

1973f90

xazhao reviewed Feb 18, 2025

View reviewed changes

dineshSajwan and others added 2 commits February 18, 2025 10:37

Merge branch 'main' into main

98f9571

implemented review comments

aa02ed8

xazhao reviewed Feb 19, 2025

View reviewed changes

dinsajwa added 2 commits February 19, 2025 11:17

implemented review comments

0fa1ae4

removed fromCfnGuardrail method from RFC scope

4745b1c

xazhao approved these changes Feb 20, 2025

View reviewed changes


		The Bedrock L2 construct simplifies the creation of multiple Bedrock features by providing a wrapper over the Bedrock L1 construct. It exposes functions that enable users to create features with minimal code. Key features include Bedrock Agent, Knowledge Base, Guardrails, Inference Profiles, and Prompt.

		For more details please refer here [Amazon Bedrock README](https://github.com/awslabs/generative-ai-cdk-constructs/blob/main/src/cdk-lib/bedrock/README.md).


		## Knowledge Base


		## Prompt management

		Amazon Bedrock provides the ability to create and save prompts using Prompt management so that you can save


		#### Specific version

		You can use the `AgentAlias` resource if you want to create an Alias for an existing Agent.

		### Create an Agent

		The following example creates an Agent with a simple instruction and default prompts that consults a Knowledge Base.


		### What are the drawbacks of this solution?

		The Knowledge Base vector stores (OpenSearch and Aurora clusters) utilize custom resource lambda functions, as there are no underlying L1 constructs available.


		### Why should we _not_ do this?

		The construct is published via the [generative-ai-cdk-constructs](https://github.com/awslabs/generative-ai-cdk-constructs/blob/main/src/cdk-lib/bedrock/README.md) repository. However, due to the increasing demands and expanding use cases, maintaining it within this repository has become challenging.

		* API Bar Raiser: @{BAR_RAISER_USER}


		The Bedrock L2 construct simplifies the creation of multiple Bedrock features by providing a wrapper over the Bedrock L1 construct. It exposes functions that enable users to create features with minimal code. Key features include Bedrock Agent, Knowledge Base, Guardrails, Inference Profiles, and Prompt.


		The resource accepts an `instruction` prop that is provided to any Bedrock Agent it is associated with so the agent can decide when to query the Knowledge Base.

		Amazon Bedrock Knowledge Bases currently only supports S3 as a data source. The `S3DataSource` resource is used to configure how the Knowledge Base handles the data source.


		Amazon Bedrock Knowledge Bases enable you to provide foundation models and agents with contextual information from your company’s private data sources. This enhances the relevance, accuracy, and customization of their responses.

		### Create a Knowledge Base

		- Amazon S3. You can either create a new data source using the `bedrock.S3DataSource(..)` class, or using the
		`kb.addS3DataSource(..)`.

		const app = new cdk.App();
		const stack = new cdk.Stack(app, 'aws-cdk-bedrock-data-sources-integ-test');


		2. `instruction` prop: Provided to associated Bedrock Agents to determine when to query the Knowledge Base

		3. embeddingsModel: Foundation model supported with bedrock.

	3. embeddingsModel: Foundation model supported with bedrock.
	3. `embeddingsModel` prop: Foundation model supported with bedrock.


		To create a vector knowledge Base, a vector index on a vector store is required. The resource accepts:

		1. `storageConfiguration` prop: An existing vector store from:

		N
		ote - you need to provide clusterIdentifier, databaseName, vpc, secret and auroraSecurityGroupId used in


		### Prepare the Agent

		he Agent constructs take an optional parameter shouldPrepareAgent to indicate that the Agent should be prepared after any updates to

Bedrock L2 Construct #688

Are you sure you want to change the base?

Bedrock L2 Construct #688

Conversation

dineshSajwan commented Jan 27, 2025

hrudu-dev left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hrudu-dev commented Feb 4, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dineshSajwan Feb 13, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dineshSajwan Feb 13, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dineshSajwan Feb 13, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dineshSajwan Feb 12, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dineshSajwan Feb 19, 2025 • edited Loading

dineshSajwan Feb 13, 2025 •

edited

Loading

dineshSajwan Feb 13, 2025 •

edited

Loading

dineshSajwan Feb 13, 2025 •

edited

Loading

dineshSajwan Feb 12, 2025 •

edited

Loading

dineshSajwan Feb 19, 2025 •

edited

Loading