#Agent架构
14 related articles
Tutorials·3 min
ReAct vs. CodeAct: Two Core Approaches to Agent Tool Calling
In-depth comparison of ReAct and CodeAct — two core Agent tool-calling architectures. From paper principles to code implementation, learn the trade-offs between reasoning+action and code execution.
Read more →
Deep Dives·3 min
Deep Dive into How Claude Code Works: A New Programming Paradigm Driven by the Agent Loop
Deep dive into how Claude Code works, including the Agentic Loop, automatic context compaction, tool calling, and permission modes that power autonomous AI programming.
Read more →
Industry Insights·2 min
Claude Code vs OpenClaw Architecture Comparison: Deep Analysis of Two AI Agent Technical Approaches
Deep comparison of Claude Code and OpenClaw AI Agent architectures—from tool governance pipelines and security sandboxes to memory systems and multi-agent collaboration.
Read more →
Deep Dives·4 min
Is Context Engineering the Core of Agent Development? A Deep Dive into Architecture and Practical Solutions
Deep dive into context engineering as the core of Agent development, covering five context modules, four pain points, and dynamic assembly solutions including compression, hybrid retrieval, multi-Agent architecture, and state machine control.
Read more →
Tutorials·4 min
Claude Code Agent Architecture Source Code Deep Dive: Agent Loop and Tool System in Practice
Deep dive into Claude Code's four core agent modules: system prompt, Agent Loop, tool system, and memory mechanism. Build a Mini Claude Code from scratch in TypeScript.
Read more →
Deep Dives·3 min
How Multi-Agent Teams Solve AI Hallucination and Make AI Reliable
Deep analysis of how multi-agent architecture solves AI hallucination. From context rot to adversarial debate mechanisms, see how Anthropic, xAI, and Kimi reduce hallucination rates from 12% to 4.2%.
Read more →
Industry Insights·2 min
Cloud AI Agent Architecture Design: Far More Than Moving Local Agents to a Server
Building cloud AI Agents requires entirely new architectural thinking. This article analyzes three core infrastructure components—durable execution platforms, execution frameworks, and dev environment tools—to help teams avoid common pitfalls when migrating from local to cloud.
Read more →
Industry Insights·4 min
Deep Dive into Claude Code's Open-Source Architecture: Six Core Design Principles and Engineering Practices Revealed
Deep analysis of Claude Code's open-source architecture: six core design principles including dual-loop mechanism, seven-step tool pipeline, four-layer token compression, multi-agent collaboration, and memory systems.
Read more →
Tutorials·2 min
How Enterprise Text-to-SQL Breaks the 90% Accuracy Barrier: A Practical Guide to Agent Architecture
A deep dive into engineering strategies for enterprise Text-to-SQL to break 90% accuracy, covering precise schema retrieval, multi-Agent architecture, self-correction, and AI coding practices.
Read more →
Industry Insights
MCP Isn't Dead! Anthropic's Two Power …
·3 min
MCP Isn't Dead! Anthropic's Two Power Moves Cut Costs by 85% + 12 Agentic Patterns Fully Explained
Deep dive into MCP's cloud repositioning strategy, how Anthropic cuts token costs by 85% with Tool Search and code sandboxes, and 12 Agentic patterns for production-grade AI Agents.
Read more →
Deep Dives·3 min
Single-Agent vs Multi-Agent: A Three-Step Decision Framework for Choosing the Right Architecture
Based on Anthropic's engineering practices, a detailed three-step decision framework for single-agent vs multi-agent architecture: bottleneck identification, technical feasibility, and business value filtering.
Read more →
Tutorials·3 min
SubAgent Architecture: The Core Implementation for Context Isolation
Deep dive into SubAgent context isolation architecture, covering parent-child Agent roles, tool definitions, run_subagent implementation, and differences from TodoList and Agent Teams.
Read more →
Tutorials·3 min
Claude Code Beginner's Guide: Installation, Cost-Saving Solutions & Source Code Architecture Analysis
Claude Code beginner tutorial covering installation, common issues, low-cost alternatives for budget users, and deep analysis of 8 Agent design patterns revealed by the 510K-line source code leak.
Read more →
Deep Dives·2 min
Agents Need Control Flow, Not More Prompts
Why do longer Prompts make AI Agents less stable? This article explains the control flow first architecture, replacing natural language control flow with code orchestration to boost multi-step reliability from 40% to over 90%.
Read more →