Search
Glossary8
Open the full glossary- Golden Set / Reference SetA golden set / reference set holds human-reviewed cases and expected standards that represent real work.
- ShellThe program that reads what you typed and goes and runs it.
- Git (version control)A tool that records how a folder changes over time: who changed what, when — and lets you go back at any point.
- Rate limitThe cap on how many requests or tokens a service lets you send in a given window; go over it and you are throttled with a 429.
- ProxyA middleman between you and the destination: your request goes to it first, it asks on your behalf, then relays the answer back.
- SFT (Supervised Fine-Tuning)SFT gives the model good input-answer pairs to imitate; a common post-training method that adjusts model weights.
- OrchestratorThe layer that decides which task goes to whom, in what order, and how results are collected; in a lead/worker pattern, the lead that delegates rather than does.
- HandoffA handoff passes a task and its necessary context from one agent to another, stating goal, progress, and boundaries.
Found 37 articles
Advanced Ontology Design: From a Reasonable Model to a Production-Grade Semantic Layer for Agents
When an ontology has to go to production: pick the right OWL profile, turn constraints into executable checks, handle identity and time, and let agents use it through a tool contract. This is about making it hold up, not making it look pretty.
Insider Threat and Access Governance: The Hardest Attack Surface Is Inside
Insider threat is not simply "having a bad actor": it covers three distinct types — malicious, negligent, and coerced — each needing different countermeasures. This article breaks down their behavioural signatures and detection logic, and why access governance beats employee surveillance.
Zero Trust Architecture: Not a Product Purchase but a Change of Assumption
In vendor decks, "zero trust" often becomes a product name; fundamentally it is only a change of assumption — no longer assuming the inside is safe. This article unpacks three core principles, four implementation layers, five common misconceptions, and why it is the same work as credential governance and insider threat management.
OkHuman Architecture Deep Dive: A Minimalist Agent Framework Written in Go
A four-layer teardown of OkHuman: 15,861 lines of dependency-free Go, one bash meta-tool, nine decoupled plugins, and three design tradeoffs that explain every architectural choice.
How We Fixed OkHuman Semantic Search and Built It a Web Search Plugin
A root-cause chain from a lying health endpoint to a one-line --pooling last fix, plus a zero-dependency search plugin with three providers and automatic fallback.
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.
Contract Clause Extraction and Risk Flagging System: From Page-by-Page Reading to Structured Risk Navigation
The most expensive thing about contract review is not the time — it is the miss. This article uses Ironclad, DocuSign, and AI together to build automatic extraction, standard comparison, and risk grading for seven categories of key clauses, so the legal team spends its time where the real differences are.
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.
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.
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.
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.
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.
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.
Automating Comparable Company Analysis: Run an Entire Sector in One Pass
The difficulty in comparable company analysis is not the arithmetic of multiples but the basis: fiscal year, currency, one-off items, and EV composition can all let two companies' numbers contaminate each other. This article breaks down the criteria for selecting peers, six categories of basis-unification rules, and the traps in choosing multiples, and gives a five-stage automation pipeline design with executable code, from data extraction to outlier handling.
What Is an API Key: It's Both Your Identity and Your Wallet
An API key is a string that tells a remote service 'this request is yours, and the bill goes to you.' This article explains why it's a secret, where it belongs (environment variables or .env — never a chat window), what to do on leakage, why it's usually wired directly to money, and a distinction most beginners miss: logging into the web version is not the same as having an API key.
Building Organizational AI Capability: From Individual Heroes to Repeatable Processes
A few people in the company being great at using AI does not mean the organization has AI capability. This article is about turning personal experience into organizational assets: capability inventory, seed users and a three-stage rollout path, consolidating personal prompts into a skill library and a template library, tiered training and documentation, and the dimensions and common failure modes for measuring whether it really landed.
Build Your First AI Workflow Without Writing Code
You can build a usable AI application with your own hands even if you cannot write code. Using a customer-support knowledge-base Q&A as the scenario and no-code platforms like Dify, Coze, and n8n, this article walks you through six steps — prepare the data, build the knowledge base, set the prompt, test, launch, monitor — each concrete enough to follow, without depending on interface details that go stale.
Resume Screening and Interview-Question Generation: Write the Criteria Down Instead of Going by Feel
The common problem in resume screening is not slowness but that the criteria live in the interviewer's head — inconsistent and hard to articulate. This article turns the hiring criteria into a document that can actually be judged, then uses Claude, NotebookLM, and Dify to extract fields, match against the criteria with the original text as evidence, and generate interview questions, while holding the line on human decision-making and compliance boundaries.
Wiring Existing SaaS into One Flow with n8n: From Manual Copy-Paste to Automation
Copying data between a CRM, spreadsheets, and messaging apps by hand is slow, lossy, and error-prone. This article uses n8n plus an LLM node to automate one process: how to inventory the manual data-moving points, pick a process, wire the nodes, and do error handling and failure alerts — and makes clear which stages should not go to the LLM and should use deterministic nodes instead.
A 30-Minute Quick Due Diligence Pass on One Company
A 30-minute quick due-diligence process you can follow as written: what to check in each of five phases, where to check it, what counts as a red flag, and what to record — closing with a one-page summary template. The goal is not an investment conclusion but judging whether a company deserves deep research.
A Reality Check on Decision Models: Why They Seem Miraculous Online but We Measured Only 54%: A Full Comparison of JEV / LAYA / KEV / CLM-8B and a Deployment Formula
In September 2026, System One decision models formed a new category within two weeks: the closed-source JEV API, then LAYA, KEV, and CLM-8B open-sourced one after another. This article does more than summarize the differences among the four; it uses six everyday scenarios to explain how they are actually used, and places the official marketing side by side with our same-question measurements: on the same set of questions, full-coverage accuracy was only 54%, but with confidence gating it reached 91.7%.
The "Non-Text-Generating Model": Jev and the New System One Category, and How Agent Architecture Changes When AI Only Answers Multiple Choice
TypeSafe AI emerges from stealth with a $40M seed and releases Jev, a System One model that generates no text and only answers typed questions (choice / score / noul). It answers multiple questions in parallel in a single forward pass, returns calibrated probabilities, with ~100ms latency, priced at $0.042/MTok.
A Complete Tutorial for Open Multi-Agent (OMA): A TypeScript Multi-Agent Orchestration Framework from Goal to Task DAG
An in-depth look at OMA (open-multi-agent) v1.8.0: a TypeScript-native multi-Agent orchestration framework with Goal-Driven Task DAG, Checkpoint resume, Consensus verification, and support for 10+ LLM providers. Includes complete code examples and a hands-on tutorial.
NVIDIA SkillSpector: An AI Agent Skill Security Scanner
GitHub Trending #11 · Open-sourced by NVIDIA · Detects 64 vulnerability patterns × 16 risk categories, scanning Agent Skills for malicious code, prompt injection, data exfiltration, and other security risks. A must-have security check before installation
200+ Skills, One Router: Engineering Practice of Agent Skill Routing
After installing 200+ skills, does the AI Agent actually get dumber? How category-by-stage matrix routing raises skill discovery from 35% to 90%, reduces incorrect tool usage by 80%, and lets any task automatically match the right skill combination.
Subagent Isolation Architecture: Reliability Lessons for AI Financial Applications, From the our HK research pipeline Data Contamination Incident to the Clean Context Design Pattern, A Complete Journey
While analyzing Stock A and Stock B back to back, Stock A's property-fund holding (carrying value HK$368 million, impairment HK$154 million — both teaching assumptions) leaked into the Stock B report. Stock B's business is consumer-goods trading and money lending; it holds no such fund at all. This is not a hallucination — it is a systemic context-contamination problem.
A Comprehensive Classification Matrix for the OpenClaw Skill Library: 227 Skills × Four Categories × a Full Business-Process Map
A complete classification matrix of the 227 installed skills in our OpenClaw system. Reorganized into four categories: everyday core tools, finance-related, coding-related, and other, with each category further arranged into 10 business stages. Includes skill usage analysis, overlap detection, and trimming recommendations.
A Lightweight Overhaul of MemoryHub in Practice: Cutting from 10 Databases to 5, and RAM from 6.2GB to 500MB
What happens when you delete Neo4j, Elasticsearch, and MongoDB from an AI memory system, keep only the Qdrant + FAISS + SQLite-vec trio, and replace the 9.1GB BGE-m3 with a 193MB bge-small-zh? A complete performance dissection, an embedding model comparison, and a trimming plan aimed at ordinary computers.
Complete Comparison of Top 10 Vector and Full-Text Databases: MemoryHub Backend Storage Selection Guide
An in-depth review of ten databases, including Qdrant, Chroma, LanceDB, SQLite-vec, FAISS, Redis, PostgreSQL (pgvector), Elasticsearch, MongoDB Atlas, and Neo4j, in AI memory system scenarios, covering performance, deployment difficulty, resource consumption, and use cases.
Complete Comparison of 15 Native Search Engines: In-Depth Analysis from Indexing Mechanisms to Application Scenarios
A comprehensive review of 15 native search engines supported by OpenClaw: indexing sources, API pricing, latency, Chinese language support, stability, and the best choice for each scenario.
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.
May 2026 W2 AI Roundup: Skills Ecosystem Explodes, Terminal Open-Source Wave, Enterprise AI Arms Race
A full record of this week's key events in the AI world: Matt Pocock open-sources 17 Agent skills and detonates GitHub, Warp terminal goes open source, DeepClaude cuts costs 17x, and OpenAI $10B vs Anthropic $1.5B face off on the same day.
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.
Three-Engine Search Strategy: DuckDuckGo + Tavily + Brave Complementary Configuration Guide
Role division, scenario selection, and cost comparison of three search engines in an Agent framework. DuckDuckGo is free and fast, Tavily provides AI summaries, Brave covers news and communities comprehensively.
Agentic Research Weekly AI Briefing: May 2026 W2
This week's AI highlights: GitHub Skills ecosystem surge, Kaltura open-sources Agent Skills, DeepClaude cuts costs by 17x, Warp terminal goes open source, GitHub Spec-Kit released. Three-engine (DuckDuckGo+Tavily+Brave) cross-validation.
Open Design: Complete Categorized Catalog of 118 Skills
Complete index of the Open Design skill system: 118 skills, 146 design systems, covering presentations, web pages, dashboards, mobile apps, video, documents, and other full-scenario use cases.