X平台推出托管式MCP，支持AI代理直连API
Related
The X platform launches a managed MCP that supports direct API connections for AI agents.
PANews reported on June 30 that, according to Cointelegraph, the X platform has launched a managed MCP (Model Context Protocol), which allows AI agents to connect Grok, Cursor, and other MCP-compatible AI tools to the X API, enabling them to access the X platform's real-time information source without any configuration.
X launches a managed version of its MCP service, allowing AI tools to access X's real-time data without any setup.
BlockBeats reported on June 30th that X announced the launch of a managed version of its MCP service, allowing AI agents to directly access real-time data from the X platform, positioning it as "the world's best source of real-time information." With the managed MCP server, developers can connect Grok, Cursor, and other MCP-compatible AI tools to the X API and X developer documentation via xurl and documentation search, quickly accessing real-time information and building AI applications without complex configuration. Currently, X provides two MCP servers for AI tools to use, aiming to further simplify the integration process between AI and the X platform, helping developers more efficiently utilize X's real-time content and development resources. Click the original link below to join the Beating · Lark AI news channel and monitor global AI hotspots and news 24/7.
Cursor introduces a Customize page for unified management of plugins, skills, and MCP.
According to Beating's monitoring, Cursor has launched a brand-new Customize page, providing unified management of plugins, skills, MCPs, sub-agents, rules, commands, and hooks at the user, team, and workspace levels. In addition to supporting the introduction of custom MCPs, the page also offers a Marketplace leaderboard showcasing the most popular tools within teams and the community; users can add them to their personal configurations with a single click. Plugins now support pre-built canvases, allowing teams to share and reuse configuration templates. The initial rollout includes the Hex canvas for data visualization and the Atlassian canvas for real-time viewing of tasks, projects, and documents. Furthermore, the Team Marketplaces, in addition to supporting local code repositories, now supports importing plugin repositories from GitLab, Bitbucket, and Azure DevOps, facilitating plugin distribution within teams.
SpaceX AI and Cursor plan to launch their first jointly developed AI model as early as Wednesday.
PANews reported on July 8th that, according to Reuters citing The Information, SpaceX AI and Cursor plan to launch their first jointly developed AI model as early as Wednesday. The report stated that the model was originally scheduled for release earlier this week but was delayed due to efficiency optimizations. The new model is expected to have rapid information processing capabilities and, in some aspects, compete with Anthropic's Opus 4.8 and OpenAI's GPT 5.5. In June, SpaceX announced its acquisition of Anysphere, the company behind the AI programming assistant Cursor, in a $60 billion all-stock transaction to strengthen its position in the AI tools market. SpaceX (formerly xAI) was acquired by SpaceX in February of this year, and SpaceX officially joined the Nasdaq 100 index this Tuesday.
Theo, an on-chain capital markets platform, invested $20 million in Fidelity's tokenized USD liquidity fund, FILQ.
PANews reported on June 30th that, according to Cointelegraph, on-chain capital markets platform Theo, through Swiss digital asset bank Sygnum, invested $20 million in Fidelity International's tokenized USD liquidity fund FILQ, becoming the first crypto-native platform to invest in the fund. FILQ is a USD liquidity fund issued on the Sygnum Desygnate platform, rated Aaa-mf by Moody's, investing in diversified short-term money market instruments. Chainlink provides on-chain net asset value and dividend data, and JPMorgan Chase (daily) audits the net asset value. Data shows that FILQ currently manages approximately $55.1 million in on-chain assets, and Theo's investment represents nearly one-third of that.
Musk: Grok 4.5 enters private testing at SpaceX and Tesla; performance may be close to or even surpass Opus.
PANews reported on June 28th that Elon Musk stated on the SpaceX platform that Grok 4.5, built on the 1.5T V9 f base model, has entered private testing within SpaceX and Tesla, with Cursor data added to supplementary training. Early evaluations show that the model's performance is close to or may even surpass Opus levels. Musk also pointed out that reinforcement learning (RL) continues to significantly improve model capabilities, and the Grok Build training framework is constantly iterating and optimizing. Furthermore, Musk revealed that new models trained entirely from scratch by SpaceX will be released monthly throughout the year.