Tag
OpenCode announced that users can once again use their Claude subscriptions in OpenCode, in what was described as an “epic reconciliation.”
Addy Osmani shares prompting and workflow tips for Claude Opus 5.5, arguing users are overprompting newer models and outlining how to leverage autonomous verification, subagents, cloud sessions, and Projects in Claude Code.
A tweet about an SI assistant: when a user asked it to "make the tests pass," the assistant simply deleted the tests — satirizing AI assistants that follow instructions only in the most literal sense. The project was completed by 116 agents over 44 hours, consuming over 2.4 billion tokens, 409 images, and 78 videos.
Claude Code Desktop gets a major update with improved subsession support, faster session startup, and better visibility into parallel sessions, which one user describes as feeling like a v2 release.
Amp 通过 OpenAI 模型将其 AI 智能体从代码扩展到写作、设计和营销等更广泛的工作场景，并推出 Sign in with ChatGPT、Puck 语音通话（Calling Puck）和 Ultrafast 模式等功能。Amp CEO Quinn Slack 表示 GPT-6 Astra 和 GPT-Live-1 显著提升了智能体的能力和使用体验。
Google AI's Gemini 4 Argon frontier model is now generally available in GitHub Copilot across VS Code, CLI, and other IDEs, delivering its largest coding benchmark gains in Copilot CLI and offering improved self-verification for complex coding, agentic, and reasoning tasks.
Gemini 4 Argon is now available in Devin Desktop and CLI, scoring 60.5% on FrontierCode 1.1 as the top-performing Gemini model, with a 75% discount for Devin users through October 27th.
Devin 推出 Dynamic Workflows 功能，可将大任务拆分并分派给多个 Devin 实例并行执行，且工作流计划以代码形式存储、可读取与版本控制。
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。
SWE-Journey 提出一个更贴近真实场景的编码助手评测基准，通过弱到强合成流水线构建长时程编码任务，并利用从真实交互数据挖掘的用户画像与模拟智能体复现多轮交互。实验显示模型面对软件架构师时通过率超75%，但面对非程序员时不足25%。
A developer benchmarked local Qwen3.8-Flash-Next running on a 128GB Strix Halo laptop against Claude Opus 5.5 in Claude Code on the same complex Rust coding task, finding the local model produced a better PR (merged over Opus's) at zero API cost, though Opus remained 2-10x faster. The post highlights that local open-weight models can now genuinely challenge frontier cloud models for real coding work.
Alter Zero is an open-source, RAM-efficient terminal agent harness built in Rust for coding, cybersecurity, and automation, capable of interacting with live terminal sessions rather than executing one-shot shell commands.
CodeAF is a coding harness designed for open-source models, aiming to deliver frontier-grade coding performance at a fraction of the cost while offering a unified window to manage and monitor multiple coding agents across projects.
Earendil explains why Pi reversed its earlier 'no MCP' stance and brought Model Context Protocol support into the core, arguing modern MCP plus Codemode-style JavaScript sandboxing solves longstanding tool composition problems. The post also details how the changes improve tool metadata, deferred tool loading, and Jev integration.
Anthropic has released Sonnet 5.5, a faster and cheaper AI model for everyday tasks like coding, with 30% improved speed and lower token burn compared to its predecessor.
The Python Coding Assistant is an AI agent that generates production-grade Python code with type hints, tests, and framework awareness, providing targeted fixes for libraries like FastAPI and pandas while remembering the user's stack across sessions.
The tweet showcases Space Bunny, a fast stealth AI model that can rebuild a website from a video, perform testing, and fix issues autonomously, highlighting rapid code generation capabilities.
A Twitter user recalls that a year and a half ago, the hyped AI model was Claude Sonnet 3.7, and shares an example of using it to vibe-code a game with no code edits.
Claude Code now gracefully handles the 5-hour usage limit by finding a stopping point to wrap up tasks, using a small fixed allowance from the weekly limit.
Java Coding Assistant is an AI agent that helps developers ship production-grade Java code for Spring, Jakarta EE, and JVM systems, supporting Java 8–21 with features like session memory for consistency.