{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"awesome-aws-kb-retrieval","slug":"aws-kb-retrieval","name":"AWS KB Retrieval","type":"mcp","url":"https://github.com/modelcontextprotocol/servers-archived/tree/main/src/aws-kb-retrieval-server","page_url":"https://unfragile.ai/aws-kb-retrieval","categories":["mcp-servers","rag-knowledge"],"tags":[],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"awesome-aws-kb-retrieval__cap_0","uri":"capability://memory.knowledge.aws.knowledge.base.retrieval.via.bedrock.agent.runtime","name":"aws knowledge base retrieval via bedrock agent runtime","description":"Enables semantic search and document retrieval from AWS Knowledge Base using the Bedrock Agent Runtime API, implementing MCP server protocol to expose KB queries as callable tools. The server translates MCP tool requests into Bedrock Agent Runtime calls, handling authentication via AWS credentials and returning structured search results with document metadata and relevance scores.","intents":["Query a company's internal knowledge base from within Claude or other MCP clients without direct AWS SDK integration","Retrieve relevant documents and context from AWS KB to augment LLM reasoning in multi-turn conversations","Integrate AWS Knowledge Base as a RAG backend into AI agent workflows without building custom retrieval infrastructure"],"best_for":["AWS-native teams already using Bedrock and Knowledge Base for document management","Enterprise organizations needing to expose proprietary knowledge bases to AI assistants via MCP","Developers building AI agents that require access to indexed company documentation"],"limitations":["Archived repository with no active maintenance — no security updates or bug fixes","Requires pre-existing AWS Knowledge Base setup; does not handle KB creation or document ingestion","Bedrock Agent Runtime API calls incur AWS charges per invocation; no local caching of results","No built-in result ranking or filtering beyond Bedrock's native relevance scoring","Limited to AWS regions where Bedrock and Knowledge Base services are available"],"requires":["AWS account with Bedrock and Knowledge Base services enabled","AWS credentials configured (IAM role or access keys with bedrock:InvokeAgent permissions)","Existing AWS Knowledge Base with indexed documents","Node.js 16+ or Python 3.8+ (depending on server implementation language)","MCP client (Claude Desktop, VS Code extension, or custom MCP host)"],"input_types":["text query strings","structured search parameters (optional filters, max results)"],"output_types":["JSON with retrieved documents","Document metadata (source, relevance score, excerpt)","Bedrock Agent Runtime response payload"],"categories":["memory-knowledge","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-aws-kb-retrieval__cap_1","uri":"capability://tool.use.integration.mcp.server.protocol.implementation.for.aws.service.exposure","name":"mcp server protocol implementation for aws service exposure","description":"Implements the Model Context Protocol (MCP) server specification to expose AWS Knowledge Base as a callable tool within MCP-compatible clients. The server handles MCP transport (stdio or HTTP), tool schema registration, request/response serialization, and error handling according to MCP specification, enabling any MCP client to discover and invoke KB retrieval without AWS SDK knowledge.","intents":["Expose AWS Knowledge Base as a discoverable tool in Claude Desktop or other MCP clients","Enable multi-tool agent orchestration where KB retrieval is one capability among many","Standardize AWS service access patterns using MCP protocol instead of custom integrations"],"best_for":["Teams standardizing on MCP for tool integration across multiple LLM clients","Organizations with existing MCP infrastructure looking to add AWS service integrations","Developers building AI agents that need consistent tool-calling semantics across providers"],"limitations":["MCP protocol overhead adds latency compared to direct SDK calls (typically 50-200ms per request)","Archived reference implementation may not reflect latest MCP spec updates","No built-in request queuing or rate limiting — relies on client-side throttling","Stdio transport limited to local execution; HTTP transport requires additional configuration"],"requires":["MCP client implementation (Claude Desktop, custom MCP host, or compatible editor)","Understanding of MCP tool schema and request/response format","Node.js 16+ or Python 3.8+ runtime","MCP SDK for the chosen language"],"input_types":["MCP tool call requests (JSON-RPC format)","Tool schema definitions (JSON)"],"output_types":["MCP tool results (JSON-RPC responses)","Tool schema metadata","Error responses with MCP-compliant error codes"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-aws-kb-retrieval__cap_2","uri":"capability://tool.use.integration.bedrock.agent.runtime.api.invocation.with.credential.management","name":"bedrock agent runtime api invocation with credential management","description":"Handles authentication and API calls to AWS Bedrock Agent Runtime service, managing AWS credentials (IAM roles, access keys, or STS tokens) and translating MCP tool requests into Bedrock-compatible invocation payloads. The server constructs agent invocation requests with query parameters, handles response parsing, and manages session state across multiple queries.","intents":["Authenticate to AWS Bedrock without exposing credentials in client code","Invoke Bedrock agents with natural language queries and receive structured results","Manage AWS API rate limits and error handling transparently"],"best_for":["Teams with existing Bedrock agent deployments seeking to expose them via MCP","Organizations using IAM-based credential management for AWS services","Developers building multi-agent systems where Bedrock agents are one component"],"limitations":["Requires valid AWS credentials with bedrock:InvokeAgent IAM permissions","Bedrock API calls incur per-invocation costs; no local caching or result deduplication","Agent response latency depends on Bedrock backend (typically 2-10 seconds)","No built-in request signing or credential rotation — relies on AWS SDK defaults","Limited to Bedrock regions where agents are deployed"],"requires":["AWS account with Bedrock service enabled","IAM credentials with bedrock:InvokeAgent and bedrock:GetAgent permissions","Bedrock agent ID and knowledge base ID configured","AWS SDK for Node.js or Python","Network access to Bedrock API endpoints"],"input_types":["text query strings","optional session context or conversation history"],"output_types":["Bedrock agent response (text, structured data, or tool calls)","Session tokens for multi-turn conversations","Error messages with AWS error codes"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-aws-kb-retrieval__cap_3","uri":"capability://data.processing.analysis.semantic.search.result.parsing.and.metadata.extraction","name":"semantic search result parsing and metadata extraction","description":"Parses Bedrock Agent Runtime responses containing Knowledge Base search results, extracting document metadata (source, relevance score, content excerpt), and reformatting results into a standardized structure for MCP clients. The server handles variable response formats from Bedrock, normalizes document references, and includes source attribution for RAG transparency.","intents":["Extract structured document metadata from Bedrock KB search responses","Provide source attribution and relevance scores to support RAG transparency","Normalize KB results into a consistent format across different Bedrock agent configurations"],"best_for":["Teams implementing RAG systems that require transparent source attribution","Developers building AI applications where document provenance is critical","Organizations needing to audit which KB documents influenced LLM outputs"],"limitations":["Parsing logic is specific to Bedrock response schema — changes to Bedrock API may break parsing","No deduplication of similar results across multiple KB queries","Relevance scores are Bedrock-native and not comparable to other search engines","Excerpt truncation may lose context for long documents","No support for custom metadata fields beyond Bedrock's standard schema"],"requires":["Bedrock Agent Runtime response with Knowledge Base search results","Understanding of Bedrock response structure and metadata fields","JSON parsing capability in the server runtime"],"input_types":["Bedrock Agent Runtime API response (JSON)"],"output_types":["Structured document list with metadata","JSON with fields: source, relevance_score, excerpt, document_id","Formatted text for display in MCP clients"],"categories":["data-processing-analysis","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-aws-kb-retrieval__cap_4","uri":"capability://memory.knowledge.multi.turn.conversation.state.management.with.knowledge.base.context","name":"multi-turn conversation state management with knowledge base context","description":"Maintains conversation history and session state across multiple KB queries, allowing clients to build multi-turn interactions where each query can reference previous results. The server manages session tokens from Bedrock Agent Runtime, preserves context across invocations, and enables follow-up queries that build on prior KB searches without re-querying the same documents.","intents":["Support multi-turn conversations where users refine KB searches based on previous results","Maintain context across multiple queries to enable follow-up questions","Reduce redundant KB queries by caching session state in Bedrock agents"],"best_for":["Interactive AI assistants where users iteratively refine knowledge base searches","Conversational agents that need to maintain context across multiple turns","Teams building RAG systems with stateful query refinement"],"limitations":["Session state is managed by Bedrock Agent Runtime, not the MCP server — limited control over context window","No explicit conversation history storage — relies on Bedrock's session management","Session tokens may expire after inactivity; no automatic refresh mechanism","Context window limited by Bedrock agent configuration (typically 4K-8K tokens)","No built-in conversation persistence across server restarts"],"requires":["Bedrock Agent Runtime with session support enabled","MCP client capable of maintaining conversation state","Session token management in the server implementation"],"input_types":["text query strings","optional session token from previous invocation","conversation history (optional)"],"output_types":["Bedrock agent response with session token","Updated conversation context","KB results with reference to previous queries"],"categories":["memory-knowledge","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":26,"verified":false,"data_access_risk":"high","permissions":["AWS account with Bedrock and Knowledge Base services enabled","AWS credentials configured (IAM role or access keys with bedrock:InvokeAgent permissions)","Existing AWS Knowledge Base with indexed documents","Node.js 16+ or Python 3.8+ (depending on server implementation language)","MCP client (Claude Desktop, VS Code extension, or custom MCP host)","MCP client implementation (Claude Desktop, custom MCP host, or compatible editor)","Understanding of MCP tool schema and request/response format","Node.js 16+ or Python 3.8+ runtime","MCP SDK for the chosen language","AWS account with Bedrock service enabled"],"failure_modes":["Archived repository with no active maintenance — no security updates or bug fixes","Requires pre-existing AWS Knowledge Base setup; does not handle KB creation or document ingestion","Bedrock Agent Runtime API calls incur AWS charges per invocation; no local caching of results","No built-in result ranking or filtering beyond Bedrock's native relevance scoring","Limited to AWS regions where Bedrock and Knowledge Base services are available","MCP protocol overhead adds latency compared to direct SDK calls (typically 50-200ms per request)","Archived reference implementation may not reflect latest MCP spec updates","No built-in request queuing or rate limiting — relies on client-side throttling","Stdio transport limited to local execution; HTTP transport requires additional configuration","Requires valid AWS credentials with bedrock:InvokeAgent IAM permissions","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.2,"ecosystem":0.49999999999999994,"match_graph":0.25,"freshness":0.52,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.15,"match_graph":0.23,"freshness":0.12}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-06-17T09:51:02.371Z","last_scraped_at":"2026-05-03T14:00:18.053Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=aws-kb-retrieval","compare_url":"https://unfragile.ai/compare?artifact=aws-kb-retrieval"}},"signature":"Yy/Ykm/+UWGb4BjEUY468Qx2eMbt1gowhzYY21SbxYCpmy9TKF9gszxDC1uoZbheEAZaW52YLvWaxJQjWXM5Aw==","signedAt":"2026-06-21T17:06:05.336Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/aws-kb-retrieval","artifact":"https://unfragile.ai/aws-kb-retrieval","verify":"https://unfragile.ai/api/v1/verify?slug=aws-kb-retrieval","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}