最新资讯

Informatica has integrated Model Context Protocol (MCP) support into its Intelligent Data Management Cloud (IDMC) platform, alongside a suite of new AI-driven features. * Informatica is among the first data management vendors to adopt MCP, an Anthropic protocol designed to enable AI models to access real-time enterprise information for enhanced accuracy. * This integration allows large language models (LLMs) to retrieve precise and up-to-date data directly from Informatica's cloud data management services. * Beyond MCP, IDMC now features expanded generative AI capabilities across data quality, governance, and master data management, leveraging its CLAIRE AI engine. * These AI advancements aim to automate data management tasks, improve data accessibility for AI applications, and enhance overall enterprise data intelligence.

The article details the process of building and deploying a Model Context Protocol (MCP) Server using Terraform on AWS. It specifically targets integration with Amazon Bedrock Agent Core to enable AI agents, such as Claude, with custom tools. - The guide covers configuring necessary AWS resources, including S3 buckets for tool definitions and IAM roles for secure access. - It demonstrates how the MCP Server provides external functionalities, allowing AI agents to perform actions by executing the defined tools. - This setup facilitates enhanced AI assistant capabilities through structured tool invocation within the Bedrock Agent Core environment.

The GitHub Model Context Protocol (MCP) Server is a tool designed to provide relevant contextual information from large codebases to AI models that consume the Model Context Protocol. It operates as a local server, processing project files and transmitting context via a Unix socket, thereby addressing AI context window limitations. The server is built for seamless integration with AI models like Anthropic's Claude, which natively supports MCP. Developers can initiate the server using the `gh mcp server` command and leverage it with `gh code assist` for various code-related tasks. It allows for providing context based on file paths, diffs, and other project elements.

Amazon Web Services (AWS) has announced the integration of the Model Context Protocol (MCP) with the Amazon Q Developer CLI, enabling serverless solutions for AI assistant interactions. * Developers can now use the Amazon Q Developer CLI to create custom tools that extend Amazon Q's capabilities, following best practices for serverless application development. * These custom tools, implemented as AWS Lambda functions, are published as MCP resources, allowing Amazon Q to discover and invoke them for dynamic problem-solving. * The new `q-dev` CLI facilitates the generation of tool definitions and the creation of serverless applications, simplifying the process of exposing backend systems to Amazon Q. * This approach leverages serverless architectures to ensure scalability, security, and cost-efficiency for AI assistant integrations that require access to external APIs or data.

Perplexity AI launched a new macOS native app, significantly enhancing its capabilities on Mac by integrating with Anthropic's Model Context Protocol (MCP). * The app leverages MCP to access and understand local system context, including files, folders, and running applications. * This integration allows Perplexity to perform system-level tasks like summarizing local documents, analyzing images, and interacting with browser content. * Users can grant specific permissions for file and application access, ensuring privacy and control. * The MCP integration represents a major step towards making AI assistants more powerful and context-aware on desktop environments.

Trail of Bits announced the development of a crucial security layer designed to enhance the Model Context Protocol (MCP). This new security layer addresses identified vulnerabilities within the existing MCP specification, particularly concerning the interactions between AI assistants (MCP Clients) and external tools (MCP Servers). Key features include robust end-to-end encryption for all exchanged context data, advanced attestation mechanisms to verify the authenticity of MCP Servers, and precise access control policies enabling granular permissions for AI assistants. The layer also integrates comprehensive audit trails to facilitate compliance monitoring and rapid incident response, aiming to secure the broader AI assistant ecosystem against malicious actors and context poisoning attacks.

The article explores a futuristic scenario where Apple Health could integrate with a Model Context Protocol (MCP). * MCP would serve as a secure interface, enabling AI assistants like a future Claude or Apple's own AI to interact with user health data. * This protocol would allow AI to process and analyze health information, offering personalized insights without direct exposure of raw data to the AI model. * The concept emphasizes stringent data privacy and security, positioning MCP as a vital standard for AI interaction with sensitive personal information. * It envisions a future where AI can provide valuable health-related assistance based on securely accessed data.

Perplexity's Mac app now supports the Model Context Protocol (MCP), a significant upgrade enhancing its capabilities as an AI assistant. MCP is an open-source framework developed by Anthropic, designed to enable AI models to interact seamlessly with external tools and resources, functioning as a "universal translator" for AI. This integration allows Perplexity to access real-time information from third-party applications, such as summarizing unread emails or finding documents in cloud storage. By tapping into a wider range of context, Perplexity can provide more accurate and relevant answers. MCP support also opens possibilities for the app to automate complex workflows across multiple applications.

Avaya is accelerating its adoption and integration of Model Context Protocol (MCP) to enhance customer experience (CX) solutions. * The MCP, developed by the Llama Index community, is designed to provide Large Language Models (LLMs) with richer, real-time contextual information from enterprise data sources. * Avaya's CX business applications, particularly for its Experience Platform, will leverage MCP to deliver more accurate and personalized AI-driven interactions. * This initiative aims to overcome the limitations of LLMs regarding real-time, accurate context, improving customer service efficiency and effectiveness. * The enhanced context capabilities will allow Avaya's AI solutions to better understand customer histories, preferences, and real-time situations, reducing agent workload and improving self-service options.

Amazon has published a guide on enhancing generative AI solutions by using Amazon Q index as an MCP Server. This integration allows Large Language Models (LLMs) to retrieve relevant information from knowledge bases for improved responses. * The article focuses on using Amazon Q index as a Model Context Protocol (MCP) server endpoint. * It demonstrates architectural patterns for integrating Retrieval Augmented Generation (RAG) with generative AI. * The solution leverages the MCP to send retrieved documents and relevant context to LLMs like Anthropic Claude. * This setup enables AI assistants to access and utilize enterprise data securely and efficiently.

Avaya has announced accelerated support for Model Context Protocol (MCP) on its Avaya Infinity Platform. * This integration is designed to significantly enhance AI assistant capabilities across Avaya's communication and collaboration solutions. * It facilitates improved context sharing for AI models, leading to the development of more intelligent virtual agents and advanced agent assist tools. * The initiative aims to boost the efficacy of AI-driven customer service interactions and streamline operational workflows. * The acceleration reflects Avaya's strategic focus on adopting cutting-edge AI protocols to foster future integrations and expand AI partnerships.

The .NET Model Context Protocol (MCP) C# SDK received a significant update, designed to simplify the creation of tools for AI models like Claude. The SDK introduces new `ToolUse` and `ToolResult` types, which streamline the parsing of tool calls and serialization of results when interacting with AI assistants. It now supports streaming of tool results, enhancing the responsiveness and efficiency of tool interactions for large outputs. Key improvements include more robust type mapping for tool arguments and easier creation of `IMCPContext` instances, making it more straightforward to integrate C# applications as tool providers for AI workflows, thereby expanding AI assistant capabilities.