#AI编程
9360 related articles
AI Agent Performance Evaluation: Two Paths and Testing in Practice
A Bilibili open class breaks down two paths of AI Agent testing: AI-empowered testing and testing AI products themselves, and explains why traditional testing fails on generative agents, plus how Python+Locust stress testing and the AIES evaluation platform divide performance evaluation work.
Doubao Work in Practice: Three Modes That Let AI Actually Get Work Done
Doubao Work practical tutorial: from chat mode to Work Mode Lite and Pro, explaining the logic for choosing among three modes with real office task demonstrations. Even complete beginners can learn to use an AI agent to handle local files, organize materials, and generate documents.
DeepSeek Harness in Practice: Integrating Browser and Phone Automation
Bilibili creator shares DeepSeek Harness in practice: integrating browser and phone automation via the EcoLite browser Skill and Android ADB wireless debugging, with a Device Workbench plugin for unified monitoring and manual takeover, expanding capability boundaries 5 to 6 times.
Stop Hand-Writing Prompts: Deconstructing the Graph Engineering Workflow for AI Agents
Deconstructing Graph Engineering for AI Agents: distinguishing control graphs, knowledge graphs, and graphs of loops, explaining the code-as-graph and model-as-graph paths, plus the Super Design team's daily bad-design triage and change-release cases with key design points.
From In-the-Loop to Out-of-the-Loop: The Next Leap in AI Agent Architecture
From in-the-loop to near-loop to out-of-loop, this article analyzes the evolution of AI agent architecture: how persistent agents, event triggers, and sandboxes form proactive agent systems, and how the four elements of context, model, prompt, and tools amplify engineers' leverage.
Building an AI Agent with Claude Code from Scratch: From Workspace to Cloud Autonomy
No coding required: build an AI Agent in four steps with the Claude Code desktop app—from workspace setup and Agent Designer to Skills, Memory, and cloud autonomy.
No Python Required: How Cursor AI Coding Is Reshaping Automated Testing and Test Development
Test engineers can do automated testing without learning Python? This article examines how Cursor and other AI coding tools empower functional testing, automation scripts, test platform development, and performance tuning, compares the cost and access restrictions of mainstream domestic and foreign tools, and offers capability transformation advice for testers.
Nearly 5,000 Affected! AI Coding Plugin Found with Backdoor, Officials Urgently Remove It
Claude Code's VS Code extension version 2.1.313 was maliciously tampered with, implanting a backdoor for remote login and persistent control, potentially affecting up to 5,000 users. Officials have removed the malicious version; this article details the attack method and cleanup steps.
Codex and Cursor Localization and Model Switching, All Solved by One Small Tool
A Bilibili creator shares CCX Plus, a small tool that solves both localization and model switching for Codex and Cursor—install and forget, reliable connections, no garbled text, immediate effect.
Dual-AI Workflow That Prevents Ship Disasters: Cloud Planning + Local Execution Across 92 Projects
A Bilibili creator shares a dual-AI workflow: cloud brain plans and reviews, local Agent executes on real machines, GitHub task ledgers bridge them—92 projects shipped, catching false test passes and license packaging failures.
Grok Bot in Practice: Lauren Tan's Methodology for Autonomous AI Code Delivery
Cursor engineer Lauren Tan on letting AI agents deliver code autonomously: verification skills and feature maps build trust, evals test skills, hard CI constraints replace review, and agents auto-merge PRs.
Generative UI Comes to Android: A2UI Lets Agents Build Native Interfaces in Real Time
A2UI is Google's open protocol that lets AI Agents generate native Android UI from declarative JSON. Learn Catalog and Surface concepts, the three-step integration flow, two-way data binding, self-correction loops, and custom component extensions.
Jetpack Glance + Remote Compose: Wear OS Tiles Evolve into Widgets
Wear OS Tiles evolve into Wear Widgets built on Jetpack Glance and Remote Compose, sharing infrastructure with phone widgets. Migration strategies, code model, and AI tools explained.
Jetpack Navigation3 Type-Safe Deep Links in Practice
A deep dive into the Jetpack Navigation3 type-safe deep link API: from DeepLinkRequest and URIDeepLinkMatcher to Kotlin Serialization parsing, lazy synthetic back stacks, and KMP sharing.
Claude's Safety Classifier Is Hurting Developers: Why a Premium Subscription Can't Retain Users
Reddit user considers canceling $200/month Claude Max subscription due to frequent safety classifier false positives on legitimate development work. Analysis of impact on paying users and AI usability challenges.
Agentic Coding with Small Models: 100x Lower Cost, Double the Success Rate
Explore how to build efficient agentic coding workflows with small language models, balancing cost, privacy, and latency. Learn about model types, local running advantages, evaluation methods, and practical tips to achieve 100x lower cost and double the success rate.
Jetpack AppState: Taking Compose State Beyond the UI Layer
Jetpack AppState extends Compose snapshot state into the data and domain layers: transform uses headless composition to derive state, and State Store enables cross-component sharing and persistence, solving Flow orchestration and process-death pain points.
Compose Performance Optimization: A Practical Guide to Outperforming Yourself
A deep dive into Jetpack Compose performance optimization: release-mode measurement, Baseline Profiles, Perfetto composition tracing, computed vs. derived state, effect API costs, and avoiding unnecessary recomposition.
Bitmap Memory Optimization in Practice: 5 Strategies to Reduce App Memory Usage
Based on Google I/O research across ~500 Android apps, this guide covers the Reduce, Reuse, Recycle Bitmap memory optimization strategies—CDN cropping, downsampling, Bitmap Region Decoder, cache deduplication, onTrimMemory handling, Play Console memory metrics, and local profiling toolchains.
TokenEconomizer: Microsoft's Open-Source Plugin Cuts Coding Agent Costs by 30.9%
Microsoft Research and UC Berkeley release TokenEconomizer, an open-source plugin that cuts coding agent token costs on SWE-Bench by 18.4% to 30.9% without modifying the agent.