Learn
74 articles. 閱讀中文版
The 74k-star Open-Source Life Guide: What Its Evidence Grading Is Worth Copying
A breakdown of a GitHub project with 74,249 stars: 772 recommendations, 2,597 primary sources, and an A/B/C evidence grading system. Compared against our T0-T5 trust model — three ideas worth borrowing, two not.
The Eight Acceptance Gates for AI-Generated Reports
Eight mechanical gates that take an AI-generated report from 'looks finished' to 'safe to deliver.' Includes real interception cases and a copyable playbook.
DeepSeek's Open-Source Super-Agent Prototype: A Source-Level Breakdown
A source-level breakdown of DeepSeek's open-source dsh-superagent-prototype: the three-plugin architecture, the event-driven core, 11 injected host services — and which designs are worth borrowing, and which are not.
PStack Deep Dive (Part 1): How a Cursor Plugin Turns Engineering Discipline into Executable Files
pstack for Cursor, unpacked: 24 named principles, 23 playbooks, and poteto-mode. A model-agnostic way to turn engineering judgment into auditable files.
PStack Deep Dive (Part 2): Two Agents, Five Special Designs, and How It Actually Lands
The executors of pstack: poteto-agent, comment-sicko, the benny automation, five special designs, and notes on landing it in OpenClaw and DeepSeek Harness.
How Databases Work: From File Cabinets to Vector Search
A beginner-to-deep guide to how databases store, index, and protect data, plus a family map of relational, key-value, document, graph, vector, and time-series engines, and how to choose.
Agent Interfaces: choosing between CLI, API, browser use, computer use, and sandboxes
Which door does an agent act through? Adapted from the MIT-licensed curriculum's Stage 8 concept chapter: eight recurring terms, a 'smallest interface first' decision table, the three tables to draw before letting an agent act (allowlist, actions, approval gates), and the four questions to ask of any browser/computer-use benchmark score. The bigger the door, the higher the risk — the first principle is to pick the smallest, most inspectable one.
Build Your First AI Agent in Seven Steps: from environment to a minimal production pass
Adapted from the MIT-licensed awesome-agentic-ai-zh integrated tutorial: one 'paper assistant' project wires up environment setup, your first LLM call, the four-part prompt, tool use, the agent loop with reflection, RAG memory, evals with observability, and the smallest interface. Every step ships minimal runnable code and safety rails; the local Ollama path costs 0 in API fees.
After You Press Enter: From 0.03 s to 0.85 s, What Your Question Goes Through
One real API call dissects the full chain of a question: 35 ms to establish the connection, microseconds to tokenize, 474 ms for the model to read your whole question, 380 ms generating it token by token — 854 ms in total. Includes the complete reproducible measurement method, and one counterintuitive conclusion: 96% of the time is not in the network but inside the model.
The AI Coding Tools Map: Recognize the Form Factor First, Memorize Names Later
Cursor, Trae, Claude Code, Codex, OpenClaw, Hermes Agent, DeepSeek Harness, WorkBuddy and Doubao Work sorted into three categories by form factor: IDE-embedded, terminal agent, and orchestration platform. For each category: what it gives you, how much control and visibility it takes away, and how people actually mix them in practice — so you can decide which category your first tool should come from.
AI Compliance and Risk Boundaries: The Legal Checklist for Enterprise Adoption
The legal homework before AI adoption is not 'find one regulation to comply with' but inventorying the risk dimensions clearly, asking the right questions, and demanding the right documents. This article gives you an inventory framework for five risk dimensions, a question list you can take straight into the legal meeting, a vendor-document request table, and the items to add to internal policies plus per-scenario management methods.
How to Cost AI: Subscriptions, Tokens, and Hidden Labor
AI adoption budgets are often severely underestimated because most people count only the subscription fees. This article teaches you to compute all three cost layers item by item — per-user monthly subscriptions, usage-billed API tokens, and the hidden labor of review and operations — and provides an estimation-sheet template you can fill in directly.
How AI Pricing Works: Why One Month Could Cost Ten Times More
What a token is, why long conversations get more expensive, what free tiers usually limit, how to choose between subscriptions and pay-as-you-go, and five concrete ways to keep your first month's cost near zero. No specific prices listed — this site only writes dollar amounts after manual verification.
Where AI Fits in the Research Process: What to Delegate and What to Keep
Breaks the investment research process into seven stages and judges one by one which can be handed to AI, which AI can only assist with, and which cannot be given up one step — with a ready-to-apply division-of-labor table and an implementation checklist that help you draw AI's capability boundary and responsibility boundary clearly.
Three Levels of AI: Chat, Collaboration, and Agent
They are all called AI, yet their capabilities and cost structures differ enormously. This article breaks the tools on the market into three levels — chat, collaboration, and agent — and gives you a decision table to first clarify which level a need belongs to, then talk tool selection and budget, so you avoid spending money on the wrong tool.
A Framework for Choosing AI Tools: Five Questions That Eliminate 90% of Options
Facing a dizzying list of AI tools, trialing them one by one is the most expensive way to choose. This article gives you five screening questions — data sensitivity, error tolerance, integration needs, usage scale, and cost of failure — each mapped to which category of tool to pick, and ends with a decision-flow table that walks you through the whole judgment.
Build or Buy: The Decision Tree for an Enterprise Agent Platform
Should we build our own Agent platform or buy one off the shelf? This article builds a decision tree from six decision dimensions — requirement uniqueness, data sensitivity, engineering capacity, maintenance cost, vendor lock-in, and compliance requirements — and gives judgment tables for the build, buy, and hybrid routes plus a total-cost-of-ownership comparison framework.
Config Files, Dotfiles, and Package Managers: What Are Those Files Starting with a Dot?
Why every tool asks you to edit a file starting with a dot, the difference between YAML and JSON, why wrong indentation breaks everything, what problems package managers like brew and apt solve, and how much regex you actually need to know.
Copying Commands from the Web into the Terminal: Why It Often Fails
You copy a command from a tutorial, paste it into your terminal, and get a syntax error or command not found — this article explains the four most common reasons: smart quotes, truncated multi-line commands, invisible whitespace characters, and pasting into the wrong shell. Each comes with an immediate verification method.
Enterprise Data Security Basics: What Data Should Never Be Fed to AI
Employees pasting customer personal data into free AI tools is one of the most common and hardest-to-prevent data leakage paths in enterprises. This article provides a four-tier data classification standard, a 'never upload' red-line checklist, a comparison of enterprise vs. consumer terms, local deployment options, and a question list for IT and legal teams.
FDE: The Person Who Delivers an Agent into an Enterprise
What problem the Frontier Deployment Engineer role actually solves: building an Agent demo and delivering one into an enterprise's real operations are two entirely different competencies. This article explains what that capability gap concretely consists of, what the Tencent Cloud ADP FDE certification examines (with verified exam parameters and official sources), and how it relates to the rest of this site.
Function Calling 101: Getting an LLM to Actually Use Tools
A clear account of the Function Calling mechanism in mainstream LLM APIs: how to write a tool JSON Schema, how the model decides whether to call a tool, how to use tool_choice for forced calls and parallel calls — with a complete runnable example, plus practical techniques for how the way you write tool descriptions decides call quality.
Git and GitHub for Beginners: Why AI Tools Keep Asking You to Clone
Why every AI programming tool's installation guide assumes you know Git. This article explains what repository, commit, branch, clone, fork, and pull request each mean, how they connect, and why non-programmers need to understand the first three.
The Evolution of the Harness: Six Eras, Each With Its New Capabilities and New Ways to Crash
From pure-prompt chatbots, CoT and ReAct reasoning prompts, and function calling — the real turning point — to MCP standardization, long-running CLI Agents, and always-on multi-Agent orchestration: era by era, what new capability the harness gained each time, and what new failure mode that capability incidentally introduced.
A Map of HK Equity Research Sources: HKEX Filings, Annual Reports, and Third-Party Data
When researching an HK-listed company, where does the information come from, which sources are trustworthy, and which can only serve as leads? This article divides HK equity research sources into five credibility tiers — statutory disclosures, company-produced information, third-party databases, market data sources, and news and research reports — and provides a 'question-to-source' mapping table.
How Machines Read Financials: From PDF to Structured Data
The first bottleneck of automating financial-statement analysis is that machines cannot read the PDF. This article breaks down the three paths — text-layer extraction, OCR, and layout analysis — explains why financial tables are especially hard to read and why human proofreading is mandatory after OCR, and gives the design concepts of structured output (JSON schema).
How the Web Talks: HTTP, Status Codes, Headers, and Invisible Protocols
Why API docs are full of status codes and headers. This article walks through a real request to clarify the four parts of HTTP — request, response, status code, header — then explains the difference between WebSocket and polling, why CORS blocks you, what a proxy does in the middle, and why SSH and private keys belong to the same family.
Install Didn't Work: Five Checks to Complete Before Reinstalling
When a tool seems unresponsive after installation, it's almost always one of five reasons: not in PATH, insufficient permissions, network blocked, version mismatch, or low disk space. This guide provides a one-minute check for each reason and explains why 'reinstalling' is rarely the solution.
Compliance Boundaries for Investment-Research AI: Disclaimers, Audit Trails and Traceability
When you put AI into an investment-research workflow, compliance is not one disclaimer line at the end of a report — it is a system-design problem: does the output constitute investment advice, are the data sources licensed, and can every conclusion be traced back to its inputs and prompt versions. This article gives a risk-dimension inventory, an audit-trail architecture, human-review checkpoint design, a question list to take to legal, and a route for strengthening internal policy.
Is My Data Safe? What Actually Leaves Your Computer When You Use AI
'Is it safe?' is too blunt a question to ever get a useful answer. This article breaks it apart: the four kinds of things that actually leave your machine when you use cloud tools, why local tools are different, what the sentence 'we don't train on your data' does and does not promise, and three rules you can actually enforce. No legal advice included.
Hallucination Risk in Finance: The Three Most Expensive Mistakes
The cost of AI hallucination in financial scenarios is severely underestimated. This article dissects the three most expensive mistakes — fabricating numbers, confusing periods and companies, and filling gaps with inference — explains what causes each one, and gives four implementable lines of defense: mandatory citation, separate statistics, low-confidence abstention, and human review.
Installation Failures on macOS: Gatekeeper, Quarantine Flags, and Missing Build Tools
When an AI-tool install stalls on a Mac, it's usually one of four things: the wrong chip's build was downloaded, Gatekeeper says it can't verify the developer, the download is flagged as coming from the internet, or the command-line build tools are missing. This article gives you one confirming action for each, in order — including the things no guide tells you, like the 'Open Anyway' button sticking around for only about an hour.
How to Read an Error Message: Anatomy, Stack Traces, and Exit Codes
Error messages aren't meant to be read word by word — they have a fixed shape: a concluding line, a stack trace, a filename with a line number, and an exit code. This article teaches why you read a stack trace bottom-up, how to find the line that actually failed, six real-shaped examples, and when to paste the whole thing into a search engine, when to give it to an AI Agent, and when to do neither.
Shareholding Topology and Double-Count Prevention: Computing Shareholdings Under Group Structures
Under a group structure, directly adding up the shareholding percentages of different layers is the most common — and most fatal — calculation error. Using a fictional teaching structure, this article demonstrates the layer-by-layer dilution computation of economic interest, voting attribution along the control chain, and the simultaneous solving of cross and circular shareholdings, and gives a method for building the shareholding structure into a queryable graph plus a double-count-prevention checkpoint list.
Before You Start: You Only Need to Know These 20 Terms
You can get started with AI Agents without writing code — as long as you understand what the docs are saying. This article splits the 20 terms that actually block beginners into four groups, gives a one-sentence definition for each plus which to read first, and explains how to use the glossary and in what order to finish a first pass.
Three-Statement Articulation and Cash-Flow Quality Checks: Spotting the Book-Profit Trap
Profit is an opinion; cash is a fact. This article takes apart the reconciliation relations of the income statement, balance sheet, and cash-flow statement, gives check dimensions such as the cash conversion ratio, working capital, and capitalized spending, adds a 'red-flag signal → account to check → verification method' cheat sheet, and demonstrates writing these checks as an automatically executable, traceable rule engine.
What an AI Agent Can Actually See on Your Computer
Which files an AI Agent can read, whether it can go online, whether it can delete things, and whether what you say gets used for training. This article splits 'what can it do' into four separately verifiable questions, and explains what permission modes, sandboxes, and working directories each block — and don't.
File Paths: Why Your File 'Cannot Be Found'
The difference between absolute and relative paths, why macOS/Linux and Windows use different separators, what the home-directory shorthand is, why names with spaces need quotes, and how to find where a file actually is on your system. The five most common path errors with their verification methods.
What Is a Terminal: The Name of Everything in That Window You Just Opened
Every install guide's first line is 'open a Terminal.' You open it, see a dark window and a line of text, and don't know where to look next. This isn't a command list — it dissects the screen in front of you line by line: what that window is, why typing can make things happen, what the prompt is telling you, and the invisible but always-active 'current directory.'
What an Agent Harness Is: The Ordinary Program Wrapped Around the Model
The horse-and-harness metaphor makes the concept clear: the layer of engineering code wrapped around the LLM — assembling context, executing tools, running the loop, managing permissions. It explains why the same model can perform worlds apart under a different harness, and why picking an AI tool is, to a large extent, picking a harness rather than a model.
What Is an AI Agent: From ChatGPT to AI That Acts on Its Own
Break down the essence of AI Agents with four components — LLM, context, tools, and loop — contrast chatbots with Agents, walk through daily, developer, and business examples plus real failure modes, and learn to judge whether your needs actually call for an Agent.
APIs, SDKs and All Those Abbreviations: What Each One Is and When You Will Need It
An API is a published agreement; an SDK is a toolkit in which someone has wrapped the details of calling that API for you. This piece spells out the difference between the two, why some tools require an API key and some do not, the distinction between REST and streaming, and whether you, as a user, ever need to write API calls yourself.
What Is Claude Code: The Reference Template for Terminal Agents
The flow of one real session explains what Claude Code is: a coding agent that lives in your terminal, reading files, editing files, running commands and iterating until the task is done. With installation methods, the CLAUDE.md memory mechanism, form factors beyond the terminal, the billing model, and the three kinds of people who should not start with it.
What Codex Is: OpenAI's Coding Agent, on Two Legs — Local and Cloud
The four forms hiding under the name Codex: the open-source Codex CLI (running in your terminal), the IDE extension, the desktop app, and Codex Web which runs tasks in parallel in the cloud. Covers installation, sign-in and billing models, sandboxing and approval mechanics, and how its form factor differs from Claude Code.
What Is Cursor: The AI Coding Tool That Lives in Your Editor
Three everyday actions — Tab completion, the Agent panel, and Plan Mode — explain what Cursor is, how it differs from a chatbot and from a terminal agent, which platforms it supports, what its billing model looks like, and the three kinds of people who should not start with Cursor.
What Is DeepSeek Harness: The Open-Source Agent Runtime Where 'Everything Is a Plugin'
DeepSeek's officially open-sourced agent runtime (npm package @deepseek-ai/dsh) bundles sessions, tools, workspaces, sandboxing, approvals, the agent loop, and dual interfaces, under the design slogan 'Everything is a Plugin'. One article covering what is in the box, what it looks like running, what it is not, and the mindset the developer-preview stage calls for.
What Is Doubao Work: The Office Agent That Operates a Virtual Desktop and Plugs Deeply Into Feishu
Doubao Work is ByteDance's AI office agent: a desktop client plus remote task assignment from your phone, an agent that actually operates inside a virtual desktop instead of only replying with text, deep integration with Feishu enterprise knowledge (meeting minutes, group chats, documents), and three modes — Fast, Expert and Task. With the billing model and the audiences it fits.
What Hermes Agent Is: An Open-Source Agent Where One Gateway Serves Twenty-Odd Messaging Platforms
Nous Research's open-source multi-channel Agent: one resident gateway process serves WeChat, WeCom, Feishu/Lark, WhatsApp, Telegram, Discord, Slack, email and twenty-odd other platforms at once, sharing one memory, generating skills automatically, and supporting natural-language scheduling plus five sandbox backends. Includes installation methods, common commands and the security defaults.
What OpenClaw Is: An Open-Source Assistant That Runs on Your Own Machine and Lives in Your Chat Apps
OpenClaw is an open-source AI assistant framework: the body runs on your own machine and appears through one Gateway in twenty-odd chat channels at once — WhatsApp, Telegram, Slack, Discord and more. One article covering its three-part architecture, the Skills ecosystem, its "models are swappable" philosophy, and the safety responsibilities you must read before using it.
What Is Trae: ByteDance's AI IDE, and the Open-Source Agent That Shares Its Name
Sorting out the three easily confused things called Trae: ByteDance's AI IDE (IDE mode and SOLO mode), the office assistant TraeWork, and the open-source trae-agent command-line framework on GitHub. With platform support, the billing model, and advice on picking between them.
What Is WorkBuddy: Tencent Cloud's AI Agent Office Workbench
WorkBuddy is the office-oriented AI agent workbench in Tencent Cloud's CodeBuddy family: assign tasks in natural language, automatic decomposition, multiple agents working in parallel, multimodal input including text, images, PDFs and spreadsheets, and delivery of office artifacts such as documents, slide decks and spreadsheets. With platform support, the Windows version restriction, and the billing model explained.
When Not to Use AI: Six Scenarios That Will Crash and Burn
The most expensive mistake in AI adoption is deploying it where it never belonged. This article covers six high-probability failure scenarios — from accuracy requirements with no human review, to rule-certain repetitive work, sensitive data, legal liability, ill-defined tasks, and volumes too small to justify a pipeline — explaining for each why it fails, what to use instead, and under what conditions to revisit the decision.
Which AI Tool Should I Use: Four Questions and You Have Your Answer
Not a comparison table — a decision tree. Ask four questions in order: what form of output you want from it, whether you are willing to touch a terminal, whether your data can go to a third party, and how long you plan to use it. Every branch points to the matching tool category and a specific article on this site.
Installation Failures on Windows: PATH, Execution Policy, and SmartScreen
When an AI-tool install stalls on Windows, the cause is almost always one of the same set: installed but the command can't be found (PATH), scripts blocked (PowerShell execution policy), or the installer blocked (SmartScreen or antivirus). This article shows you how to confirm whether the program actually got installed, where it put its files, and the one confirming action for each kind of block.
Your First AI Task: How to Give Instructions It Gets Right
Why first tasks usually fail — not because the tool is weak, but because the instruction has no output, no input, and no destination. Three elements, three real examples (each with its vague version for contrast), and how to judge how much freedom to give.
Harness Engineering: The Workbench Around Your Agent
Your agent writes code in demos but breaks in production because you trained a brain without building a workbench. Five diagrams and everyday metaphors unpack Harness Engineering core concepts and rollout steps.
What Is LangGraph: From Linear Chains to State-Machine Agents
Using kitchen and subway-map metaphors plus five diagrams, this guide explains LangGraph states, nodes, and edges, with a complete runnable example.
What Is Vibe Coding: Fast, but Fragile
Vibe coding turns plain-language prompts into a working prototype in minutes. So why does it collapse in production? Three failure modes and five guardrails, so you can move fast without breaking things.
Complete LangChain Tutorial 2026: Building Enterprise-Grade LLM Applications from Scratch
The most detailed LangChain 2026 tutorial: covering LCEL, RAG, Agent, LangGraph, LangSmith. From installation to production deployment, with complete code examples.
MemoryHub v2.0 System Architecture In-Depth Analysis: From Capture Daemon to MCP Real-Time Memory Capture
A complete breakdown of MemoryHub v2.0's dual-mode capture engine, Qdrant vector storage, MCP four-platform integration, Dashboard real-time monitoring, three-layer deduplication, and ten-year lifecycle design.
May 2026 LLM API Pricing Landscape: Complete Comparison of DeepSeek, Qwen, GLM, Kimi, MiniMax, and Doubao
Cross-verified through multiple search engines including Exa, Tavily, and Brave, this compiles per-million-token pricing for DeepSeek V4, Qwen3.6-plus, GLM-5, Kimi K2.5, MiniMax M2.7, Volcano Engine Doubao Seed 2.0, and Claude/GPT, with a Sub2API cost setup guide.
Cross-Channel Memory Hub: A Full Record of the Memory System Architecture Design for OpenClaw Agent
From daily forgetting to automatic cross-channel integration: a complete record of the architectural evolution of our AI assistant memory system, Session isolation issues, time zone boundary bugs, the three-layer protection mechanism, and the Cron automation integration pipeline.
OpenClaw Creator and Infrastructure Automation: From Full Podcast Workflow to Self-Healing Servers
An in-depth breakdown of the creative tools in Awesome OpenClaw Use Cases (Podcast, YouTube, AI video editing, game development) and infrastructure DevOps solutions (self-healing servers, n8n orchestration), with complete workflow architecture diagrams.
OpenClaw Productivity Revolution: 22 Automation Solutions Reshaping Daily Workflows
From Personal CRM to Second Brain, from multi-channel customer service to autonomous project management, a deep dive into the 22 most practical productivity automation solutions in Awesome OpenClaw Use Cases, with architecture analysis and practical recommendations.
A Deep Dive into Search Channels: The Underlying Principles and Division of Labor Strategies of DuckDuckGo, Tavily, and Brave
A complete technical primer on three search channels: their respective indexing mechanisms, ranking algorithms, return formats, and cost models, plus how to divide work precisely within AI Agent workflows.
In-Depth Analysis of Addy Osmani's agent-skills Open Source Project
A complete introduction to Google Chrome Engineering Director Addy Osmani's agent-skills project: the design philosophy behind 21 AI-assisted development skills, real-world applications, and how it inspired the upgrade of our slash command system.
Agent Framework Capability Matrix: Hermes → OpenClaw → Claude Code Full Capability Overview
One matrix to understand the full capability distribution of the three-layer Agent framework: search API, LLM backend, MCP protocol, skill system, role division, communication channels.
AI Security and Red Teaming: Prompt Injection, Jailbreak Defense
A full landscape of security threats to AI systems: Prompt Injection, Jailbreak, Data Poisoning, and practical defense strategies and red team testing methods.
LLM Evaluation and Selection Guide: How to Scientifically Choose Models in 2026
A systematic LLM selection framework: Benchmark interpretation, real-world scenario testing methods, cost/performance matrix, DeepSeek V4 vs Qwen vs Llama comparison.
MCP (Model Context Protocol) Ecosystem Explained: The Three Core Primitives of Tool, Resource, and Prompt
A complete guide to Anthropic's MCP protocol: the design principles behind Tools, Resources, and Prompt Templates, practical configuration, and Claude Code integration.
Prompt Engineering Systematic Methodology: From Few-shot to ReAct
A complete methodology for Prompt Engineering: principles and practical applications of techniques such as Few-shot, Chain-of-Thought, Tree-of-Thought, ReAct, and Self-Consistency.
RAG In-Depth Principles and Practice: A Complete Guide from Chunking to Rerank
The complete tech stack for Retrieval-Augmented Generation (RAG): Embedding selection, Chunking strategies, vector database comparison, Rerank optimization, with practical code examples.
Three-Layer Agent Collaboration Framework: Hermes → OpenClaw → Claude Code Architecture Explained
An in-depth analysis of a three-layer AI Agent collaboration architecture built on Hermes Agent, OpenClaw, and Claude Code: the responsibilities of each layer, the communication mechanism, and why it is designed this way.
Understanding the Transformer Architecture: Attention Is All You Need
An in-depth understanding of the core components of Transformer: Self-Attention, Multi-Head Attention, Positional Encoding, and how they replaced RNNs to become the foundational architecture of modern LLMs.