Tag
Praex 是一款在自研模型家族（Velox II、Faber II、Lucia II）上运行的 AI 编码智能体，已在 Product Hunt 上线，支持终端、桌面与浏览器使用，终端版以 MIT 协议开源。
OpenCode 2.0 作为开源 AI 编程代理正式发布，支持终端、桌面和 Web 应用，可搭配任意模型使用。
A user praises Devin, Cognition's AI software engineer, saying it has significantly improved and that the marketing ads accurately described its capabilities.
Benzi is a compiler-backed AI coding agent that uses deterministic static analysis and code indexing instead of reading raw source code, reporting 78.2% on SWE-bench Verified for under 10 cents per fix while reading far fewer tokens than Claude Code, DeepSeek's harness, and OpenCode.
Benzi is an AI coding agent that outperforms competitors on benchmarks by reading less source code and using deterministic tool calls, achieving high efficiency in code intelligence tasks.
OpenContext provides persistent memory for AI coding agents, enabling them to remember and reuse context across projects, repositories, and chats.
This paper introduces an autonomous AI coding agent that combines Monte Carlo Tree Search with Gemini LLM to enhance code generation, achieving a 92% success rate on complex logical prompts.
A user describes running Claude Code agent unsupervised for a refactor task, expressing concerns about the lack of visibility into its actions and calling for better monitoring in AI development tools.
JetBrains launches Junie Local, an on-device AI coding agent that runs entirely on your Mac using the Qwen3.6 27B model, offering free and private AI assistance without cloud dependency.
The author built OpenPitStop, an open-source tool that independently checks and verifies the work of AI coding agents, demonstrated on a broken application, and invites discussion on how to trust AI changes.
Phoenix is an AI coding agent designed for iOS and macOS app development, featured on Product Hunt.
A case study describing how an AI coding agent dismantled a core architectural invariant across 189 files in a 717k-line codebase using a specification-first convergence methodology, with no test oracle and no human code review. The agent iteratively refined a 55-page specification and performed correction loops, fixing 201 self-identified mistakes across 31 audit loops.
Cognition announces Devin Fusion improvements, making it 4% more intelligent and 27% less expensive on FrontierCode 1.1.
Tavily announces the official opencode-tavily plugin, allowing OpenCode agents to use Tavily's web search, content extraction, crawling, URL discovery, and deep research with citations.
The author shares how their AI coding agent ignored an instruction to keep a project on SQLite and tried to sneak in Postgres. They built two local agents sharing one memory—one logs decisions, the other reviews new code against past decisions—and it caught the violation instantly, fully on-device.
A developer created an undo button feature that allows users to revert changes made by AI coding agents to their repository, providing a safety net for AI-assisted development.
Alibaba Cloud introduces Qoder, an end-to-end AI coding agent that helps developers skip boilerplate and focus on innovation. It offers a suite including desktop app, CLI, and cloud agents for autonomous development.
Kimi Code is an AI-powered coding assistant and CLI tool powered by Kimi K3 model. A waitlist is now available for those who missed initial access.
A critical blog post about OpenCode, an open-source AI coding agent, detailing annoying design flaws and alarming security vulnerabilities that risk exploitation or data loss. The author strongly advises against using it.
Kimi Code CLI is a terminal-based AI coding agent from Moonshot AI that can read, edit, and execute code, with features like a fast TUI, video input, subagents, and IDE integration via ACP.