#AI编程
9368 related articles
How Can AI Product Managers Make Agents Truly Understand the Business? An Analysis of Agent Skills in Practice
Why do general-purpose agents always fail to understand the business? This article explains how Agent Skills packages industry experts' processes and standards into Skill.md manuals, enabling agents to execute tasks according to professional frameworks, improving result reliability and consistency while reducing debugging and development costs.
Chrome DevTools Adds WebMCP Support: AI Tool Debugging Made Clear
Chrome DevTools now supports WebMCP: view available tools in the Application panel, trace call inputs/outputs, manually trigger tools, and jump from records to tool source code.
Open-Source Coding Agent fx Comes to T3 Code: Full Support for Web, Desktop, and Mobile
Open-source coding agent fx is now integrated into T3 Code, supporting web, desktop, and mobile. T3 Code serves as a coding agent control plane with Supervised and Auto modes, compatible with existing models and skills, and simple installation and configuration.
GitHub Octoverse 2026: AI Agent Engineering Drives Explosive Growth in Commit Volume
GitHub Octoverse 2026 reports 220M monthly merged PRs, 7.38B commits, and 30.3M new repos, with agentic engineering driving growth. We unpack the methodology shifts and the changing development paradigm behind the numbers.
CarMaker AI Agent: A New Paradigm for Simulation Engineering Efficiency
IPG engineer Zhang Li demonstrates how CarMaker AI Agent uses MCP tools to restructure simulation workflows—from test objective definition to automated execution and result comparison—letting engineers focus on objective setting and result review while reducing time spent on repetitive configuration and parameter searching.
WorkBuddy Step-by-Step Tutorial: Master AI Office Automation from Scratch in 7 Lessons
Complete guide to the WorkBuddy AI office tutorial: 7 lessons take you from zero to AI Agent proficiency, covering proposals, spreadsheets, data analysis, posters, software development, and browser automation, plus capability boundaries and target audience analysis.
Foreign Trade Worker Builds GreenShield Decryption Assistant: F8 One-Key Application + AI Agent Auto-Approval
A foreign trade worker built an open-source tool with AI Agent: F8 one-key GreenShield decryption request, wrapped as an MCP Server so AI Agents can auto-request, wait for approval, and continue working—without bypassing the official approval chain.
GitHub Universe 2026: The Ambition of Full-Stack Evolution
GitHub Universe 2026 aims to build the world's most comprehensive end-to-end software delivery system. This article examines its full-stack evolution strategy, positioning shift, layered integration logic, impact on developer workflows, and the opportunities and uncertainties of platform concentration.
Octoverse 2026: Developers Rebuild Programming with AI Agents
GitHub Octoverse 2026 reports record commits and pushes, with AI agents as the key growth driver. This analysis explores how agent programming impacts developer workflows, code authorship, and repository activity metrics.
Building an MCP Server from Scratch: Let AI Apps Call Your Agent Tools
A hands-on guide to building an MCP server from scratch so AI apps like Claude Desktop can call your Agent tools, covering version negotiation, tool discovery, invocation, error handling, and testing.
Claude Code in Practice: Mastering AI Coding Agents, MCP, and Skills
A deep dive into Claude Code: slash commands, claude.md context budget, four permission modes and Hooks security, MCP servers vs. Agent Skills, and Anthropic's TDD autonomous workflow.
The AI Agent Era: How Engineers Can Stay Sharp and Keep Control of Their Code
How can engineers stay sharp and keep control of their code when AI agents assist development? Based on a YouTube creator's views, this article explores strategies like continuous learning, active review, and system understanding, while facing open questions in rapid evolution.
A Complete Guide to High-Quality LoRA Training
A systematic end-to-end guide to high-quality LoRA training, covering dataset construction and cleaning, base model and hyperparameter selection, the regularization image debate, checkpoint evaluation, and iteration strategies.
Codex Builds a JianYing Video Pipeline Automatically: A Zero-Basics Fully Automated Editing Walkthrough
A Bilibili tech video explains how to use Codex with Python to write JSON drafts directly and automatically build a JianYing video pipeline: from script, TTS voiceover, and subtitles to fully automated timeline assembly, including a zero-basics hands-on guide to environment deployment, MVP testing, and troubleshooting ghost drafts.
The AI Game Development Inflection Point: Vibe-Coded Games Enter the Playable Era
Simon Willison argues October 2026 is the vibe game development inflection point: Claude Opus 5.5 breaks through in 3D modeling, music generation, and game design, with models making genuinely fun games for the first time.
AI Testing Is Outdated? The 2026 Shift Toward AI Evaluation and AI-Led Testing
In 2026, AI testing is shifting from assisted testing to AI-led testing and AI evaluation. This article explains why evaluation replaces assertions, the dual-version pinning principle, the three-layer evaluation path for LLMs/knowledge bases/agents, and tool selection advice including Promptfoo.
Hand-Building an AI Virtual Girlfriend: A Practical Guide to Assembling a 2D Agent
Bilibili creator 沐野楓園 hand-builds an AI virtual girlfriend "Teacher Loki": DeepSeek as the brain, GPT-SoVITS for voice, character cards for soul, runs on a 3060, starting at five or six yuan. This article breaks down the technical architecture, cost control, and anthropomorphic tuning methods.
The Right Way for Beginners to Learn AI: Don't Get the Four-Stage Learning Path Wrong
Stuck at the door learning AI? A Bilibili creator splits the AI learning path into four stages: Python/PyTorch/math foundations, hands-on deep learning, choosing a direction for a complete project, and reading papers plus coding to break into large model deployment. Check which level you're stuck at.
Building Enterprise-Grade AI Agents from Scratch: A Three-Stage Learning Path from Bilibili's Comprehensive Tutorial
Analysis of Bilibili's 748-episode AI Agent beginner course: cognition section clarifies Agent vs. LLM differences, advanced covers LangChain, AutoGen, and multi-Agent collaboration, hands-on implements intelligent customer service and automated data analysis projects, with source code, courseware, and a learning roadmap.
AI Web Novel Writing from Scratch: A Complete Guide to Downloading General Agents and Drag-and-Drop Toolkit Import
A detailed guide to AI web novel writing from scratch: downloading four general Agents (Doubao, Workbody, DeepSeekHonest, Codex), drag-and-drop toolkit import into the dialog box, using one sentence to have the Agent auto-unzip and explain usage, and model configuration tips for letting another Agent fill in API details.