#AI编程
9716 related articles
Alibaba Open-Sources page-agent: Add an AI Operator to Any Website with One Line of Code
Alibaba's open-source page-agent embeds an AI operator into any webpage with one script tag — no Python, no headless browser, supports 60+ LLMs, uses DOM structure instead of screenshots for lightweight, controllable web automation.
Harness AI Engineering in Practice: Enterprise-Grade Development Methodology with Claude Code and Codex
From Vibe Coding to enterprise-grade AI engineering: how Claude Code, Codex, SuperPower, and SDD methodology power real-world development at scale.
Anthropic Releases Claude Opus 5.5: Faster, Cheaper, and Record-Breaking Alignment
Anthropic launches Claude Opus 5.5: 40% cheaper and 30% faster than Opus 5, with record alignment scores and improved agentic coding capabilities.
Genie One MCP Goes Generally Available: What It Means for the AI Integration Ecosystem
Genie One MCP is now Generally Available, offering enterprises a unified AI agent connectivity and governance layer built on the Model Context Protocol (MCP).
llm-typesafe: Integrating TypeSafe AI's Jev Judgment Model into the LLM CLI Tool
Simon Willison's llm-typesafe plugin brings TypeSafe AI's Jev model to the LLM CLI, enabling type-safe structured outputs for classification and scoring tasks.
Claude Code v2.1.280 Update Breakdown: Opus 5.5 Becomes the Default Model
Claude Code v2.1.280 launches with Claude Opus 5.5 as default (1M context, $0.20/M cache reads), major stability fixes, symlink security hardening, and expanded VSCode integration.
Using AI Agents to Optimize Rust Code: A New Approach to Performance Engineering
Explore a new approach to Rust performance optimization: using AI Agents to iteratively improve code speed with benchmark-driven feedback loops. Analysis of feasibility, key practices, and risks.
dsh-agent-studio Goes Open Source: Visually Configure AI Agent Prompts, Tools, and Skills
dsh-agent-studio is an open-source visual plugin for configuring AI Agents. Freely assemble prompts, tools, MCP, Skills, and fallback models, with per-sub-agent model and thinking intensity settings.
DeepSeek V4.1 Flash Deep Dive: The Architectural Revolution Behind Its Rock-Bottom Pricing
DeepSeek V4.1 Flash: How CED architecture, CSA2 sparse attention, and FP4 caching compress KV Cache by 437x, cut costs 60%, and beat expensive flagships on agentic benchmarks.
DeepSeek V4.1 Flash Hands-On: 65 Million Tokens Burned for Under $1
DeepSeek V4.1 Flash: 65M tokens in 89 minutes for under $1. Deep dive into its CED architecture, sub-1KB KV cache, Ingram conditional memory, and MIT license.
DeepSeek's Triple Move, Xiaomi MiMo V2.6 Tops Open-Source Rankings, Grok 4.7 Closes In on Opus
DeepSeek releases DSEC sandbox, pivots to Huawei chips, hires CFO; Xiaomi MiMo V2.6 tops open-source intelligence index; Grok 4.7 closes in on Opus 5 on coding benchmarks.
Firedrill: An Open-Source Framework for Safely Testing AI Agents with Stateful Synthetic Tools
Firedrill is an open-source framework for safely testing AI Agents using stateful synthetic tools — no production accounts needed. Supports fault injection, permission changes, call auditing, and CI/CD integration across 30+ tools including Gmail, Stripe, and HubSpot.
Claude Code Beginner's Guide: From Installation to Git Workflow in Practice
A beginner's guide to Claude Code: core capabilities, CLI & IDE plugin setup, switching to domestic LLMs via CC Switch, and hands-on Git workflow using pure conversation.
OpenCode Getting Started Guide: Complete Installation and Configuration Walkthrough
A complete OpenCode tutorial covering desktop and WSL installation, model configuration, rule files, Agent types, custom commands, MCP services, and Agent SQL support.
What Is Codex? A Complete Installation Guide for OpenAI's Next-Gen AI Coding Agent
Codex is OpenAI's next-gen AI coding agent. This guide covers what Codex is, its 4 forms, comparison with Claude Code, and full installation steps for Git, Node.js, VS Code, and the main app.
Claude Code Beginner's Complete Guide: Master the AI Coding Workflow in 28 Minutes
A complete beginner's guide to Claude Code — covering installation, project building, code editing, debugging, terminal commands, CLAUDE.md rules, and agent mode, plus the 5 most common mistakes to avoid.
The Right Approach to Agent-Native Apps: Shared Action Layers Over UI Puppeteering
BuilderIO's Agent-Native framework advocates for a shared action layer over UI puppeteering, letting UI, agents, HTTP, MCP, A2A, and CLI share one unified action contract.
WeWeb MCP: Let AI Agents Build Your App While You Stay in Control
WeWeb MCP lets AI agents like Claude Code and Cursor auto-build app pages, workflows, and data models — all changes land in a visual editor for human review and control.
vgpu: A WebGPU TypeScript Library Designed for AI Agents
vgpu is a WebGPU TypeScript library built for AI agents, featuring typed shader imports, a minimal GPU-first API, and unified browser/Node/test suite execution.
Grok 4.7 Released: The Flagship Model for Coding and Knowledge Work
Grok 4.7 launches as xAI's strongest model for coding and knowledge work, offering longer reasoning, better self-verification, and improved safety calibration at the same price as Grok 4.6.