Show HN: Coding Agent Survey – Which coding agents do you use?

- This is a community survey (posted on Hacker News) asking developers which AI-powered coding agents they use — tools that can autonomously write, debug, or refactor code based on natural-language prompts. - Coding agents differ from traditional autocomplete (e.g., GitHub Copilot) or chat-based assistants: they can plan, execute multi-step edits, run terminal commands, and self-correct without direct human keystrokes. - The term "agent" here follows the current AI trend where a system is given a goal and a set of tools (file editor, terminal, browser) and works independently until done. - Notable tools in this space include: Cursor (IDE with deep agent mode), Windsurf, Devin (a fully autonomous "AI software engineer"), Claude’s Code/Cline, and open-source frameworks like OpenHands (formerly OpenDevin) and Aider. - The survey matters because the coding-agent category is evolving fast — there is no clear market leader yet, and usage patterns (e.g., which tasks people trust agents with, which models they use, how often they intervene) will shape how these tools are built next.