Amazon Bedrock KB MCP. Ground LLM answers in your private AWS data.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Amazon Bedrock KB connects your AI agent directly to AWS Bedrock Knowledge Bases. It lets your agent run managed RAG, semantic searches, and sync vector data sources inside AWS.
You get direct access to massive corporate datasets—no custom vector pipelines needed. It's for agents that need grounded, verifiable answers from your own AWS data.
What your AI agents can do
Get knowledge base
Retrieves details for a specific AWS Bedrock knowledge base.
List data sources
Lists all AWS storage buckets bound to a specific Bedrock Knowledge Base.
List ingestion jobs
Checks the status of explicit AWS Bedrock Knowledge Base sync operations.
The agent generates LLM responses using internal documents stored in the Knowledge Base.
The agent queries vector indexes to pull the top-K text chunks and the source document URLs.
The agent checks which AWS storage buckets are currently feeding the Knowledge Base.
The agent tracks the real-time status of document chunking and syncing operations.
The agent lists all configured Knowledge Bases within your AWS region.
The agent lists the specific vector stores and embedding models available to the service.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Amazon Bedrock KB MCP Server: 6 Tools for Data Operations
These tools let your agent manage the entire lifecycle of your Knowledge Base, from listing sources to running complex retrieval queries.
019d755aget knowledge base
Retrieves details for a specific AWS Bedrock knowledge base.
019d755alist data sources
Lists all AWS storage buckets bound to a specific Bedrock Knowledge Base.
019d755alist ingestion jobs
Checks the status of explicit AWS Bedrock Knowledge Base sync operations.
019d755alist knowledge bases
Lists all available AWS Bedrock knowledge bases in your region.
019d755aretrieve
Sends a query to a vector index and retrieves the top-K text chunks securely via AWS Bedrock.
019d755aretrieve and generate
Generates LLM responses that are explicitly grounded in content from the Bedrock Knowledge Base.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Amazon Bedrock KB, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Amazon Bedrock KB connects your AI agent straight to AWS Bedrock Knowledge Bases. It lets your agent run managed RAG, semantic searches, and sync vector data sources right inside AWS. You get direct access to massive corporate datasets—no custom vector pipelines needed. It's for agents that need grounded, verifiable answers from your own AWS data.
get_knowledge_base lets your agent pull the details for a specific AWS Bedrock knowledge base. list_knowledge_bases shows you all the knowledge bases set up in your region. list_data_sources tells you which AWS storage buckets are feeding a specific knowledge base. list_ingestion_jobs checks the live status of document syncing and chunking operations. retrieve sends a query to a vector index and pulls the top-K text chunks and their source document URLs securely via AWS Bedrock. retrieve_and_generate generates an LLM response that's explicitly grounded in content from the Bedrock Knowledge Base.
Your agent can generate grounded answers by using retrieve_and_generate, which builds the LLM response using internal documents stored in the Knowledge Base. It queries vector indexes using retrieve to pull the top-K text chunks and their original document URLs. You can see exactly what documents feed the knowledge base by calling list_data_sources.
You can track the real-time syncing status of chunking and syncing operations with list_ingestion_jobs. You can view all configured knowledge bases in your AWS region using list_knowledge_bases. You can also get details on a single knowledge base using get_knowledge_base.
How Amazon Bedrock KB MCP Works
- 1 First, subscribe to the server and provide your AWS IAM Role/User Access Credentials.
- 2 Next, your AI client uses the tools to query the Bedrock KB. For instance, calling
list_data_sourcesconfirms the active document inputs. - 3 The agent gets back a grounded answer or a list of sources. This output is then fed back into the conversation to solve the original query.
The bottom line is, you plug your agent into AWS and let it do the hard work of data retrieval and context grounding.
Who Is Amazon Bedrock KB MCP For?
This is for developers and architects who need to run complex, data-intensive queries against private, enterprise data without managing underlying data infrastructure. If your job involves synthesizing answers from huge, decentralized document sets, this tool is for you.
Builds RAG workflows quickly. They use this to connect their agent to AWS without having to host or maintain vector databases or chunking logic themselves.
Audits the system. They use this to check ingestion status and confirm that the origin documents are mapped correctly from the chat interface.
Prototyping context-grounded queries. They use this to instantly test query accuracy against exact data chunks before deploying a full model.
What Changes When You Connect
- Managed RAG: You don't build the vector pipeline. Calling
retrieve_and_generatelets your agent generate answers grounded directly in your internal documents. - Source Visibility: Need to know where the answer came from? Using
retrievepulls exact top-K text chunks and their original S3 document URLs. No guesswork. - Audit Trail: Before trusting the data, run
list_data_sourcesto confirm exactly which AWS buckets are feeding the KB. You see the source, not just the answer. - Operational Insight: Use
list_ingestion_jobsto check the sync status. You know if the data is fresh, or if the chunking pipeline failed. - Discovery: Start with
list_knowledge_basesto see every KB you've set up. It's a simple way to audit your entire Bedrock setup. - Foundation Models: This bypasses the need to manage your own vector database, letting you focus on the prompt, not the infrastructure.
Real-World Use Cases
Compliance Check: What is the policy for X?
A compliance officer needs to know the current policy for Project Phoenix. They ask their agent, which then runs list_data_sources to confirm the policy manual is connected. The agent then calls retrieve_and_generate, which pulls the relevant section and answers with the exact policy citation.
Debugging Data Sync: Why is the KB missing the latest manual?
A data scientist notices a policy change isn't reflected. They run list_ingestion_jobs to check the sync status. The tool shows the job failed, pointing them directly to the chunking pipeline error they need to fix.
Researching a Topic: Show me the top 3 quotes on 'quantum computing' from our docs.
A researcher needs specific, unedited snippets. They use the retrieve tool, which queries the vector index and returns the top 3 text chunks, along with the document and page numbers, allowing for perfect citation.
System Audit: What KBs and models are active?
A cloud architect needs a quick inventory. They run list_knowledge_bases to see all active KBs, and then use list_knowledge_bases to confirm the assigned embedding models for compliance.
The Tradeoffs
Assuming the answer is always in the KB.
The user asks a general question and expects the agent to answer, but the agent fails because the context is too broad or the source is outdated.
→
First, check the sources. Use list_data_sources to verify the relevant AWS buckets are connected. If you need to know the status of the data, run list_ingestion_jobs before asking the question.
Manually piping raw data into the prompt.
The user copies and pastes a 50-page PDF's text into the chat box to ask a question. This is slow, messy, and hits context window limits.
→
Let the tool do the work. Use retrieve_and_generate. This tool handles the data ingestion and context window limits automatically, using the live AWS data stream.
Not confirming model availability.
The agent fails to initialize because the wrong embedding model is referenced, leading to a vague error about context failure.
→
Run list_knowledge_bases first. This shows the exact embedding models assigned to the KBs, ensuring your agent uses the correct, supported model.
When It Fits, When It Doesn't
Use this server if your core problem is synthesizing answers from massive, private documents already stored in AWS S3/Bedrock. It's ideal when you need verifiable context and granular control over data sources. Don't use it if your data lives outside of AWS, or if you only need a simple chatbot interface without source citations—in that case, a general-purpose messaging tool will suffice. If you only need to check the status of a single bucket, you can call list_data_sources directly, but for full RAG, you need the whole suite.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Amazon Bedrock. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 6 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Manually sifting through PDFs and SharePoint sites is a time sink.
Today, answering a specific question means opening five different tabs: the compliance manual, the HR policy PDF, the engineering spec sheet, and the last week's meeting transcript. You copy-paste snippets, you cross-reference version numbers, and you spend half your time trying to figure out which document is the most current truth.
With the Amazon Bedrock KB MCP Server, your agent queries all those sources in one go. You don't copy, paste, or cross-reference. You just ask, and the agent returns a single, cited answer derived from the entire corpus.
Amazon Bedrock KB MCP Server: Get answers, not just documents.
The biggest time saver is the elimination of the manual data audit. You don't have to check the sync logs or manually verify the source buckets. The server handles the connection and status tracking for you.
What's different now is that the system guarantees the context is real-time and traceable back to the source. You get the insight immediately, without the manual verification steps.
Common Questions About Amazon Bedrock KB MCP
How do I check if my Knowledge Base is connected to the right data sources using list_data_sources? +
The list_data_sources tool shows every AWS storage bucket currently bound to the Knowledge Base. This confirms the scope of data the agent can access.
What is the difference between retrieve and retrieve_and_generate? +
retrieve only pulls the raw, top-K text chunks from the vector index. retrieve_and_generate uses those chunks to write a final, synthesized answer, which is usually what you actually want.
How do I know if the data has been updated? Should I use list_ingestion_jobs? +
Yes. The list_ingestion_jobs tool provides the real-time status of the syncing process. It tells you if new documents have been chunked and mapped without errors.
Can I see what KBs are available in my AWS account? Use list_knowledge_bases. +
Yes. Running list_knowledge_bases gives you a full list of every Knowledge Base configured in your AWS region, allowing you to audit your entire setup.
If I need to find out which knowledge bases I have access to, should I use list_knowledge_bases? +
Yes, use list_knowledge_bases. This tool displays all available AWS Bedrock knowledge bases mapped to your region, letting you select the correct context for your queries.
How do I see the sync status of my data buckets using list_ingestion_jobs? +
list_ingestion_jobs lists the sync operations for your attached data sources. It shows if the chunking pipeline completed successfully and how many documents were processed.
Can I query the vector index using the `retrieve` tool? +
The retrieve tool queries the vector index directly. It pulls the exact top-K text chunks and their source document URLs without generating an answer.
What is the function of `retrieve_and_generate`? +
retrieve_and_generate performs two steps: it retrieves relevant context and then generates an LLM response grounded in that specific data. It provides the answer and the source snippets.
Can my AI agent directly run RAG without calling external LLMs? +
Yes! Use the retrieve_and_generate capability. Your agent passes the query and a designated Bedrock model ARN. Bedrock handles fetching chunks from the local vector index and synthesizing the final answer inside AWS boundaries, returning a fully grounded response instantly.
How can I check if new uploaded documents are successfully indexed in my agent? +
Just ask your agent to list ingestion jobs for a specific Knowledge Base ID and Data Source ID. It will report back the exact status (e.g., SYNCING, COMPLETED, FAILED) of chunks being mapped to your vector layout.
Can I see exactly where an answer came from in my documentation? +
Absolutely. Both the standard retrieve functionality and retrieve_and_generate calls will parse out the specific origin document URLs (e.g., S3 paths) and expose the exact raw text snippets that mathematically matched your query vector.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
OOPSpam
Block spam submissions on your forms and comments with an AI-powered filter that catches bots without annoying real users.
HubSpot Operations Hub
Manage workflows, CRM properties, pipelines, owners, and property groups through natural conversation.
Worksection
Collaborative project management to manage tasks, projects, and timers with AI using Admin API v2.
You might also like
Farcaster (Decentralized Social Protocol)
Interact with the Farcaster protocol — manage channels, moderate casts, and explore social graphs directly from your AI agent.
Zakeke
Manage customized product designs, 3D configurations, and orders via Zakeke.
UKG Ready
Manage employee lifecycle, HCM, payroll, and time via UKG Ready.
