MCP+AI Agent:人工智能应用新框架

Odaily星球日报Published on 2025-05-13Last updated on 2025-05-13

Abstract

MCP协议有望成为推动下一代AI Agent的重要引擎。

一、MCP 概念介绍

此前人工智能领域,传统的聊天机器人多依赖于通用的对话模型,缺乏个性化的角色设定,导致其回应往往显得单一且缺乏人情味。为了解决这一问题,开发者引入了“人设”概念,即为 AI 赋予特定的角色、性格和语气,以使其回应更贴近用户的期望。然而,即使 AI 具备了丰富的“人设”,它仍然仅仅是一个被动的回应者,无法主动执行任务或进行复杂的操作。因此开源项目 Auto-GPT 应运而生。Auto-GPT 允许开发者为 AI 定义一系列工具和函数,并将这些工具注册到系统中。当用户提出请求时,Auto-GPT 会根据预设的规则和工具,生成相应的操作指令,自动执行任务并返回结果。这种方式使得 AI 从一个被动的对话者,转变为一个主动的任务 AI。

尽管 Auto-GPT 在一定程度上实现了 AI 的自主执行,但其仍面临着工具调用格式不统一、跨平台兼容性差等问题。为了解决这些问题 MCP(Model Context Protocol,模型上下文协议)应运而生,其旨在解决 AI 在开发过程中面临的主要挑战,特别是在与外部工具集成时的复杂性。MCP 的核心目标是简化 AI 与外部工具的交互方式,通过提供统一的通信标准,使得 AI 能够轻松调用各种外部服务。传统上,要使大规模模型执行复杂任务(如查询天气或访问网页),开发者需要编写大量的代码和工具说明,这极大地提高了开发的难度和时间成本。而 MCP 协议通过定义标准化的接口和通信规范,显著简化了这一过程,允许 AI 模型更快速、有效地与外部工具进行交互。

MCP+AI Agent:人工智能应用新框架

二、MCP 与 AI Agent 的融合

MCP 与加密 AI Agent 之间属于相辅相成的关系,两者之间的区别在于 AI Agent 主要聚焦于区块链的自动化操作、智能合约执行和加密资产管理,强调隐私保护和去中心化应用的集成。MCP 更侧重于简化 AI Agent 与外部系统的交互,提供标准化协议和上下文管理,增强了跨平台的互操作性和灵活性。加密的 AI Agent 可以通过 MCP 协议进行更高效的跨平台集成和操作,从而提升其执行能力。

此前的 AI Agent 具备一定的执行能力,例如通过智能合约执行交易、管理钱包等。然而这些功能通常是预定义的,缺乏灵活性和适应性。MCP 的核心价值在于为 AI Agent 与外部工具(包括区块链数据、智能合约、链下服务等)的交互提供了统一的通信标准。这种标准化解决了传统开发中接口碎片化的问题,使得 AI Agent 能够无缝对接多链数据和工具,并大幅强化 AI Agent 的自主执行能力。例如,DeFi 类 AI Agent 通过 MCP 可以实时的获取市场数据并自动优化投资组合。此外,MCP 为 AI Agent 打开了全新的方向,即多个 AI Agent 协作:通过 MCP,AI Agent 可以按职能分工协作,组合完成链上数据分析、市场预测、风控管理等复杂任务,提升整体效率与可靠性。链上交易自动化:MCP 串联各类交易与风控 Agent,解决交易中滑点、交易磨损、MEV 等问题,实现更安全、高效的链上资产管理。

三、相关项目

1.DeMCP

DeMCP 是一个去中心化的 MCP 网络。它致力于为 AI Agent 提供自研开源的 MCP 服务,为 MCP 开发者提供商业收益共享的部署平台,并实现主流大型语言模型 (LLM) 的一站式接入。开发人员可以通过支持稳定币(USDT、USDC)来获取服务。截止 5 月 8 日,其代币 DMCP 市值约$ 1.62 M。

MCP+AI Agent:人工智能应用新框架

2.DARK

DARK 是一个在可信执行环境 (TEE) 下的 MCP 网络,基于 Solana 构建。代币$DARK 上线币安 Alpha,截止 5 月 8 日市值约 1181 万美金。目前,DARK 的首款应用程序正在开发阶段,它将通过 TEE 和 MCP 协议为 AI Agent 提供高效的工具集成能力,让开发者能够通过简单的配置快速接入多种工具和外部服务。尽管产品尚未全面发布,用户可以通过邮箱候补的方式加入早期体验阶段,参与测试并提供反馈。

3.Cookie.fun

Cookie.fun 是一个专注于Web3生态中 AIAgent 的平台,旨在为用户提供一个全面的 AI Agent 指数和分析工具。该平台通过展示 AI Agent 的心智影响力、智能跟随能力、用户互动和链上数据等指标,帮助用户了解和评估不同 AI Agent 的表现。4 月 24 日 Cookie.API 1.0 更新推出了专属 MCP 服务器,该服务器包含即插即用的智能体专用 MCP 服务器、专为开发者和非技术人员设计、无需任何配置。

MCP+AI Agent:人工智能应用新框架

信息来源:X

4.SkyAI

SkyAI 是基于 BNB Chain 构建的Web3数据基础设施项目,旨在通过扩展 MCP 来构建区块链原生的 AI 基础架构。该平台为基于Web3的 AI 应用提供可扩展且可互操作的数据协议,计划通过集成多链数据访问、AI 代理部署和协议级实用程序,简化开发流程,从而推动 AI 在区块链环境中的实际应用。目前,SkyAI 支持来自 BNB Chain 和 Solana 的聚合数据集,数据量已超过 100 亿行,未来还将上线支持以太坊主网和 Base 链的 MCP 数据服务器。其代币 SkyAI 上线币安 Alpha,截止 5 月 8 日市值约 4270 万美金。

 四、未来发展

MCP 协议作为 AI 与区块链融合的新叙事在提升数据交互效率、降低开发成本、增强安全性和隐私保护等方面展现了的巨大潜力,尤其在去中心化金融等场景中,具备广泛的应用前景。然而,当前大多数基于 MCP 的项目仍处于概念验证阶段,尚未推出成熟的产品,导致其代币价格在上线后出现持续下跌,例如 DeMCP 代币在上线不到一个月的时间里价格已下跌 74% 。这一现象反映出市场对 MCP 项目的信任危机,主要源于漫长的产品开发周期和缺乏实际应用落地。因此,如何加快产品的开发进度、确保代币与实际产品之间的紧密关联,以及提升用户体验,将是当前 MCP 项目面临的核心问题。此外,MCP 协议在加密生态系统中的推广仍面临技术整合的挑战。由于不同区块链和 DApp 之间的智能合约逻辑和数据结构存在差异,统一标准化的 MCP 服务器仍需投入大量开发资源。

尽管面临上述挑战,MCP 协议本身仍展现出巨大的市场发展潜力。随着 AI 技术的不断进步和 MCP 协议的逐步成熟,未来有望在 DeFi、DAO 等领域实现更广泛的应用。例如,AI 代理可以通过 MCP 协议实时获取链上数据,执行自动化交易,提升市场分析的效率和准确性。此外,MCP 协议的去中心化特性,有望为 AI 模型提供一个透明、可追溯的运行平台,推动 AI 资产的去中心化和资产化进程。MCP 协议作为 AI 与区块链融合的重要辅助力量,随着技术的不断成熟和应用场景的拓展,MCP 协议有望成为推动下一代 AI Agent 的重要引擎。然而,要实现这一愿景仍需解决技术整合、安全性、用户体验等多方面的挑战。

风险提示:

所提供的信息仅供参考,不应被视为购买、出售或持有任何金融资产的建议。所有信息均真诚提供。然而,我们不对此类信息的准确性、充分性、有效性、可靠性、可用性或完整性作出任何明示或暗示的陈述或保证。

所有加密货币投资(包括收益)本质上都具有高度投机性,并涉及重大损失风险。过去、假设或模拟的表现并不一定代表未来的结果。数字货币的价值可能上涨或下跌,购买、出售、持有或交易数字货币可能存在重大风险。您应该根据个人投资目标、财务状况和风险承受能力,仔细考虑交易或持有数字货币是否适合您。BitMart 不提供任何投资、法律或税务建议。

Trending Cryptos

Related Reads

Bitmine Increases Holdings to 5.7 Million ETH, Included in Russell 1000 Index, Tom Lee Says Only One Step Away from '5% Target'

Bitmine Increases Ethereum Holdings, Nears 5% Supply Target and Joins Russell 1000 Bitmine Immersion Technologies (BMNR) acquired an additional 27,084 ETH last week, bringing its total holdings to 5,700,040 ETH, which represents 4.7% of Ethereum's total supply. This puts the company at 94% of its goal to own 5% of ETH's circulating supply. Despite a challenging week where ETH price fell 8% and BMNR's stock dropped 13%, Bitmine continues its accumulation strategy, with Chairman Tom Lee attributing the market weakness to quarter-end "window dressing" by investors. In a significant development, Bitmine was added to the Russell 1000 large-cap index on June 26. Lee expects this inclusion to attract hundreds of institutional investors, as passive funds and ETFs rebalance to match the index. The company's total assets, including crypto, cash, and securities, stand at $9.8 billion. Of its ETH holdings, 4,879,157 are currently staked, generating an estimated annualized revenue of $211 million. Lee projects this could rise to $246 million once all ETH is staked through its MAVAN network. Bitmine remains the world's largest corporate Ethereum treasury and the second-largest overall corporate crypto treasury, behind Strategy's substantial Bitcoin holdings. Lee reaffirmed the company's commitment to steady buying throughout 2026, expressing belief that the crypto market is still in the early stages of a spring cycle. At the time of writing, ETH was trading around $1,565, while BMNR stock was near $13.56, down over 90% from its 52-week high.

marsbit17m ago

Bitmine Increases Holdings to 5.7 Million ETH, Included in Russell 1000 Index, Tom Lee Says Only One Step Away from '5% Target'

marsbit17m ago

AGI Countdown: OpenAI's Chief Research Officer Makes Major Statement — The Window for Humanity is 'Very Small'

The countdown to AGI has begun, according to OpenAI's Chief Scientist Mark Chen, who states the window for human-centric progress is "very small." Chen argues that AI is reaching a point where models can perform "self-sustaining research," autonomously driving innovation in fields from mathematics to programming. He points to the proliferation of AI's "superhuman" insights—akin to AlphaGo's legendary "Move 37"—across disciplines as evidence of this shift. Chen firmly dismisses claims that scaling laws are plateauing or that pre-training is dead, asserting the field remains on an exponential curve. He cites OpenAI's successful bet on reasoning models like o1 as proof that fundamental breakthroughs are still possible. The future of research, he suggests, lies with "Vibe Researchers"—humans who provide high-level direction and "taste" while AI handles execution and orchestration of complex, long-horizon tasks. However, significant hurdles remain. Chen highlights a "benchmarking crisis," where models can overfit to existing tests without gaining true generalization. He also notes the "jagged frontier" of AI capabilities, where systems excel at advanced reasoning but struggle with contextual, continual learning from everyday experiences. Despite these challenges, he expresses confidence that these gaps will be closed. In a personal reflection, Chen shares that post-AGI, his wish is to open a noodle shop—a metaphor emphasizing that when AI masters knowledge and innovation, uniquely human experiences, warmth, and storytelling will become the ultimate form of value.

marsbit32m ago

AGI Countdown: OpenAI's Chief Research Officer Makes Major Statement — The Window for Humanity is 'Very Small'

marsbit32m ago

China's No.1, Closing in on OpenAI, Mysterious "Sweeping Monk" Rises to Top Seven Globally

A mysterious Chinese AI project named "MopMonk" (meaning "Sweeping Monk") has achieved a top-ranking result on the globally recognized CyberGym cybersecurity benchmark. With a 73.1% success rate, it ranks seventh worldwide and first among Chinese entries, performing closely behind OpenAI. The significance lies in the benchmark itself. CyberGym, created by UC Berkeley, is considered a premier "Olympics" for AI security. It tests models on over 1500 real-world software vulnerabilities, requiring them to not just identify but actually generate working exploits (PoCs) in a complex, offline environment. This moves beyond simple knowledge to testing an AI's practical "execution" capabilities. MopMonk's approach is notable. It uses the open-source MiniMax M3 model from Shanghai as its powerful reasoning "brain," leveraging its strong coding skills and long context window. However, the key to its performance is a custom-built, multi-agent security framework—its "Harness." This system uses structured "vulnerability memory" to efficiently guide the search for exploits, allowing multiple agents to explore in parallel while sharing lessons learned from failures. This engineering layer effectively translates the model's intelligence into actionable, iterative testing steps. The project remains highly secretive, with no official website or team information, embodying the "dark horse" spirit of its literary namesake. Its success highlights a potential industry shift: beyond simply scaling model size, the engineering of specialized agent systems (the Harness) is becoming a critical differentiator for real-world AI application performance, especially in complex domains like cybersecurity.

marsbit1h ago

China's No.1, Closing in on OpenAI, Mysterious "Sweeping Monk" Rises to Top Seven Globally

marsbit1h ago

Will History Repeat Itself? Fidelity Lists Five Catalysts to End the Crypto Winter

Fidelity's new report suggests that the current crypto winter for Bitcoin may be nearing its end, identifying five potential catalysts that could drive a market turnaround based on historical patterns. First, Bitcoin's approximately four-year cycle, driven by its halving mechanism, historically marks peaks and troughs. The last bottom was in November 2022, potentially pointing to the next around November 2026, though cycle length can vary. Second, clearer regulation has often preceded past bull markets. The focus is now on the CLARITY Act, which aims to clarify US digital asset oversight between the SEC and CFTC. Its passage could unlock domestic activity currently held back by legal uncertainty. Third, Federal Reserve monetary policy plays a role. A shift to lower interest rates tends to correlate with rising crypto prices by reducing borrowing costs and boosting risk appetite, though markets may price this in well ahead of any official change. Fourth, the emergence of breakthrough applications can fuel investor interest. Current trends like real-world asset tokenization, AI-related crypto infrastructure, and stablecoins are being watched, but history shows the biggest catalysts are often unexpected. Fifth, a new wave of institutional adoption could be a trigger. While ongoing adoption in 2026 hasn't sparked a new bull run, a major unexpected move—like a significant purchase by a tech giant or adoption as a hedge in a global crisis—could create a powerful new narrative. Fidelity concludes that while the market is in a downturn, historical turning points have often resulted from a combination of such factors, and the next phase for Bitcoin may depend on which of these catalysts materializes first.

Foresight News1h ago

Will History Repeat Itself? Fidelity Lists Five Catalysts to End the Crypto Winter

Foresight News1h ago

Trading

Spot

Hot Articles

Discussions

Welcome to the HTX Community. Here, you can stay informed about the latest platform developments and gain access to professional market insights. Users' opinions on the price of AI (AI) are presented below.

活动图片