Tag
Cursor data shows top developers (P99) produce 46 times more lines of code than median developers. Olivia Moore predicts that AI will empower those previously limited by old systems, enabling them to become top performers in every role.
A discussion about an incident where an AI coding agent deleted a production database, arguing that blaming the developer is misplaced when proper review processes are not in place.
An opinion piece examines whether AI coding tools like Claude Code and Copilot truly enhance developer skills or merely accelerate flawed decision-making, highlighting the need for new metrics to evaluate human-AI collaboration in engineering.
A tweet highlights that P99 developers using Cursor AI produce 46x more lines of code than the median developer, suggesting AI empowers those previously constrained by older systems.
A tweet announces a game to test approving terminal commands against auto-review, with Cursor announcing the new auto-review mode for safer execution.
Cursor has released an auto-review mode that allows agents to execute tool calls with fewer approval prompts while maintaining safety.
A data-driven report from Cursor analyzing developer habits, revealing trends such as doubled coding speed year-over-year, larger PRs, and the growing adoption of agentic coding.
Cursor AI released the Cursor Developer Habits Report, sharing findings on how software development is changing based on a comprehensive dataset on AI coding from all model families.
A GitHub repository containing official Cursor plugins for developer tools, including agent workflows, code review, documentation, and CI integration.
Cursor has shipped a new 'thermo-nuclear' code review feature for aggressive AI code review. Matt Pocock reviews its effectiveness.
A tweet introduces a workflow where GPT-5.5 xhigh plans and delegates implementation to Composer 2.5 subagents via the pi-cursor-sdk, claiming it outperforms using either model alone. The linked GitHub repo is an open-source SDK that integrates Cursor models into the pi agent runtime.
Cursor's head of AI engineering shares a 14-minute free resource on how to build autonomous coding agents that ship production code, revealing architectural insights that a $1,000 vibe-coding course won't teach.
A commentary contrasts a Chinese trader's $2,000 course sales based on impressive setups with an engineer's high AI adoption rate that caused 9 of 11 production incidents, highlighting the gap between perceived and actual performance with AI tools.
Cursor's engineering notes reveal that agent failures often stem from the harness (scaffolding) rather than the model itself, with different tool formats across providers causing silent errors and reliability issues.
CodeGraph builds a local knowledge graph for AI coding agents to index codebase relationships, cutting token usage by ~59% and execution time by ~49% compared to traditional search methods.
Cursor shares key lessons from building cloud agents, emphasizing that providing a full development environment is critical for agent output quality, and that long-running agents require durable execution and enterprise-like infrastructure.
Cursor's annualized revenue reached $3 billion with over 3,000 large customers, and SpaceX has an option to acquire the company for $60 billion upon its public listing on June 12.
Cursor shared the training methods for its self-developed programming model Composer 2, including large-scale continuous pre-training, long-range reinforcement learning, and an internal benchmark CursorBench, which brings the model's programming performance to a top level.
Cursor is doubling usage for new users invited on the Teams plan for the next month, as announced by Michael Truell.
Composer 2.5 achieves 63.2% on CursorBench at $0.55 per task, nearly matching top models at 20x lower cost.