Articles from cursor.com

Cursor released a native iOS app in public beta that lets developers launch and control AI coding agents from their phone, with cloud agent support and remote control.

Cursor's Design Mode lets users point, draw, or narrate UI changes in the browser, with agents editing the underlying code in real time.

Coder productivity has doubled year-over-year, with AI-generated code surviving review at higher rates, while top 1% of users far outpace median developers.

Cursor's cloud agents perform best when given full development environments, durable execution via Temporal, and a harness that shifts control to the agent.

Composer 2.5, a major update to Cursor's AI coding assistant, improves long-horizon task performance via targeted RL feedback and synthetic data, built on Kimi K2.5.

Cursor launches cloud agent development environments with multi-repo support, Dockerfile-based config as code, and per-environment governance controls.

Cursor improves its coding agent by iterating on context management, evaluation metrics, and model-specific harness customization, treating the harness as a software product.

Cursor introduces a TypeScript SDK for building programmatic agents using its runtime, harness, and models, available in public beta.

A University of Chicago study of 500 companies found AI usage rose 44% as models improved, with developers shifting to more complex and cross-system tasks.

Cursor uses real-time reinforcement learning on live user interactions to ship improved Composer model checkpoints every five hours.

Cursor is building local indexes for regular expression search to reduce agent wait times from 15-second ripgrep runs in large monorepos.

Cursor released Composer 2, a frontier-level coding model achieving strong benchmarks (CursorBench 61.3, Terminal-Bench 61.7) at competitive pricing.

Cursor trains Composer to generate its own compact summaries mid-task, reducing context errors by 50% while using 80% fewer tokens than traditional prompting.

Cursor uses a hybrid online-offline eval system, CursorBench, built from real developer sessions to better distinguish model quality than public benchmarks.

Cursor adds over 30 new plugins from partners including Atlassian and Datadog to extend its AI-powered coding agent's capabilities across infrastructure and productivity tools.

Cursor Automations let developers create always-on agents triggered by events (Slack, Linear, GitHub, PagerDuty, webhooks) to automate code review, monitoring, and maintenance tasks.

Cursor's support team uses Cursor with MCP servers to collapse code, logs, and team knowledge, achieving 5–10x throughput gains.

Cursor releases cloud agents that operate in isolated VMs, enabling autonomous code testing, PR creation, and UI verification without local resource conflicts.

Cursor built OS-level sandboxes for coding agents, reducing approval prompts by 40% via Seatbelt on macOS, Landlock+seccomp on Linux, and WSL2 on Windows.

Cursor built a multi-agent harness that recursively delegates coding tasks across planners and workers, achieving ~1000 commits per hour on a browser research project with minimal human intervention.

Cursor cuts time-to-first-query from hours to seconds on large codebases by securely reusing a teammate's prebuilt index via Merkle tree-based syncing and access proofs.

Cursor improved Bugbot's bug-finding rate by defining a custom AI-driven resolution rate metric that enabled systematic hill-climbing experiments, raising resolution from 52% to over 70% and doubling resolved bugs per PR to roughly 0.5.

Cursor's multi-agent system, using planners and workers, scaled to run hundreds of autonomous coding agents for weeks on projects like building a browser, finding simpler coordination beats complex distributed-computing approaches.

Cursor's official guide details best practices for coding agents, emphasizing planning, context management, rules, and iterative review to maximize productivity.

Cursor introduces dynamic context discovery, pulling only necessary data into the context window via files, reducing agent tokens by up to 46.9%.

Cursor introduces Debug Mode, an agent loop that generates hypotheses, instruments code with runtime logs, and requests human verification to reliably fix bugs that stump standard AI models.