Loading activity…
@cursor_ai
Coding agent for building ambitious software
Technology
BleepingComputer reports the modified extension was available through Cursor's marketplace for about 11 hours on October 29.
AI
Cursor says its marketplace's modified 2.1.313 build attempted a persistent SSH backdoor and affected at most 5,000 users.
AI
Cursor says agents can click and type outside the browser, using the iOS Simulator to test features they build.
XAI
Cursor says users can request dashboards or reports while agents work, then share finished canvases with teammates by link.
XAI
Cursor says Muse Wonder scores 46.1% on CursorBench at default effort, the highest of any Muse model.
AI
Cursor says Gemini 4 Argon scores 52.8% on CursorBench at default effort, on par with Opus 5.5.
AI
Cursor says follow-up questions in the same chat can produce new charts with /visualize.
AI
Anthropic says it costs around 75% less to run on average than Haiku 4.5 and improves coding and computer use.
AI
Cursor says users can check in on agents, send replies and start new tasks from the app.
XAI
Cursor says subagents report back in a follow-up turn, with stream() and wait() continuing until they all finish.
AI
Cursor says run.steer() adds a message to the next turn, while a mid-task subagent keeps working in the background.
XAI
Cursor also says GLM 5.3 Max is the best-scoring open-weight model on CursorBench 4.0.
AI
Cursor says it spent time with Innate's team to capture what building means to them. It quotes co-founder Vignesh Anand: “The seven of us here do the work of 50.”
XAI
Cursor says its /visualize command can analyze data and show answers inline in the Agents Window.
AI
Anthropic says Sonnet 5.5 runs more than 30% faster than Sonnet 5. In pre-release max-effort tests, Artificial Analysis measured per-task costs roughly 50% higher than Sonnet 5’s.
AI
Cursor credits tighter prompts, selective tool loading, better caching and compressed file reads for the savings.
Technology
Anthropic just dropped Claude Fable 5.1, a powerhouse AI model that doubles down on tough coding challenges and long projects, making complex work feel more manageable.
AI
E2B integration lets regulated teams run Cursor cloud agents on controlled machines.
AI
Official account shares update on the AI code editor integration.
AI
Official post from the AI code editor's account highlights the model's benchmark performance.
