Knowledge Base

Archify has the agent write a typed JSON IR and lets a deterministic compiler turn it into a self-contained HTML artifact. I read the repository end to end and built four diagrams with it, watching what the validator refuses, why deliver, visual-check, and human review are kept apart, and what changes when this blog’s own mermaid is ported over.

I booted DeepSeek Harness with a single npx command and no API key, dumped its 145-row boot tree with --dump-config, and edited the patch layer by hand. The web profile ships with every model-facing tool switched off, and the reason turns out to be interesting. I also walked straight into the trap where patching a row config silently kills a CLI flag.

WebMCP is a W3C Community Group proposal in which a web page registers JavaScript functions and HTML forms as "tools" on document.modelContext, and the browser mediates those tools to the user's agent. This article lays out the imperative and declarative APIs and the boundaries the browser guarantees, straight from the spec, then calls real tools with Chrome 152 and agent-browser 0.36 to see how it differs from accessibility-tree automation.

What CO₂e means, how much CO₂ has built up in the air, how much of the 1.5 °C budget is left, and how big a tonne really is, shown in four 3D visuals instead of paragraphs.

DeepSeek Harness (dsh) turns the model adapter, the tool registry, the session log, and even the agent loop into Cordis plugins, then assembles the product by stacking patch layers on top of an empty tree. Two weeks after release it passed 200k GitHub stars. This is a source-level analysis of its kernel (Cordis), profile and bundle composition, the rule that the session log is the source of truth, capability seams, self-modification tools, and the way it runs Claude Code and Codex as subagents.

"Rails is slow" is an unanswerable question, because a Rails app and a Go app are never doing the same work. So I narrowed it: inside one runtime, what do the framework and the ORM cost? Four runtimes, each built twice, all returning identical JSON from the identical four SQL statements, measured three times on Docker Linux with MySQL. The magic tax in CPU per request came out at ×7.67 for Ruby, ×6.14 for Python, ×2.62 for Node and ×2.51 for Go — and I counted, object by object, exactly what Active Record builds on every request.

bitchat is a P2P messenger with no accounts and no servers. Nearby devices form a BLE mesh; distant peers are reached over Nostr. The center of this piece is the four-layer store-and-forward stack that answers "the recipient is not here right now" — a sealed outbox, courier envelopes addressed by nothing but an HMAC tag plus spray-and-wait, GCS-filter public history sync, and Nostr mailboxes — traced through the code.

Hallmark is Together AI’s anti-AI-slop design skill. It ships almost no executable code — its entire intelligence lives in 106 markdown files. This piece dissects the SKILL.md router and its progressive context loading, the narrowing decision pipeline, and — at its center — the 58 slop-test gates, contrasting it with skill-injecting projects like Superpowers and SkillSpector.

Paper notes on The Llama 3 Herd of Models: 15T tokens, a 405B flagship, 128K context, and a design philosophy of 'managing complexity' (dense over MoE, SFT → rejection sampling → DPO instead of PPO) — the paper where an open-weight model first reaches GPT-4-class performance. It completes Llama 2's skeleton through scale, data, and simplicity.

Paper notes on Llama 2 covering its commercial license, a base model scaled to 2T tokens with 4k context and GQA, and the Llama 2-Chat alignment pipeline (SFT → two reward models → iterative RLHF with rejection sampling + PPO → GAtt). It fuses LLaMA's open base lineage with InstructGPT's RLHF at production quality.

Floci is an always-free, open-source local AWS emulator with no account, no auth token, and no feature gates. A single Quarkus + GraalVM native container emulates 68 AWS services behind one port, 4566. We analyze its Smithy-spec protocol routing, descriptor-driven service catalog, four storage modes, and real Docker execution — against LocalStack.

Paper notes on LLaMA covering its core strategy — train smaller models for longer on publicly available data, optimized for inference — the result where a 13B model beats 175B GPT-3, the RMSNorm/SwiGLU/RoPE architecture changes, the 2048-A100 training scale, and how it kicked off the open LLM ecosystem.

Firecrawl is a data API that turns any website into clean, LLM-ready Markdown or structured JSON. It orchestrates in TypeScript, cleans in Rust, converts to Markdown in Go, races several scrape engines in a waterfall, and drives crawls with a hand-built Postgres queue. We analyze the scrape/crawl/map/search/extract pipelines and contrast them with Browser Use as "two ways to feed the web to LLMs."

SkillSpector is NVIDIA's security scanner for AI agent skills. It vets a skill before install to find prompt injection, data exfiltration, and malicious code. We analyze its structure — a LangGraph map-reduce graph that fans out to 25 analyzers (static patterns, AST, taint, YARA, MCP, LLM semantics) and reduces them to a single risk score — against Superpowers and ponytail, which inject skills.

ponytail packs a single discipline — the "lazy senior dev" laziness ladder — into one SKILL.md and ships it to 16 agents via every mechanism (skill, hook, command, MCP, plugin), proving a ~54% code cut with a real agentic benchmark. We analyze it against Superpowers, and the idea that a skill is a bundle of discipline, distribution, and measurement rather than code.

Browser Use is a Python agent in which an LLM drives a real browser. It pre-chews the page into an indexed list of interactive elements for the LLM, drives the browser via CDP instead of Playwright, and runs the session with an event bus and watchdogs. We look at why this fits LLMs better, picking up from the earlier Playwright analysis.

Cline is a coding agent that lets you approve every action inside VS Code and roll back at any time. It also detaches its core from the host via a protobuf/gRPC boundary (ProtoBus + HostBridge), so the same core runs in the editor and in a terminal CLI. We analyze Plan/Act approval, checkpoints, and the host abstraction, contrasting them with OpenCode's headless engine.

OpenCode externalizes model metadata to models.dev and even hand-rolls its own LLM protocol layer, so any provider attaches with a single line of data. A single Effect-based HTTP engine is shared by the TUI, web, desktop, Slack, and editors (ACP), while two generations — legacy and V2 — coexist. We analyze the structure against Qwen Code's single-vendor platform.

Paper notes on InstructGPT covering its core method — the three-step RLHF recipe (SFT → reward model → PPO) — the alignment result where a 1.3B model beats 175B GPT-3, gains in truthfulness and toxicity, the alignment tax and PPO-ptx, and the limitations.

Paper notes on GPT-3 covering its core ideas: 175B scaling, in-context learning (zero/one/few-shot), weighted-sampling training data, headline benchmark numbers, and the data-contamination and bias limitations.

A retrospective on building a macOS desktop app (Tauri + Rust + React) with an AI agent. The decision to make markdown files — not the database — the source of truth, the research I did before writing any code, the features I chose not to build, and the macOS permission bug that took the longest to crack. A story about judgment more than code.

Semble is a Python library that splits code into chunks with tree-sitter, fuses Model2Vec static embeddings with BM25 via RRF, and applies code-aware reranking — delivering millisecond code search on CPU alone. Where CodeGraph solves the same problem with an AST knowledge graph, Semble solves it through retrieval. This post analyzes the architecture by contrasting the two approaches.

CodeGraph is a TypeScript tool that uses tree-sitter to parse source code, builds a local SQLite knowledge graph of symbols, edges, and files (with FTS5), and exposes that graph via an MCP server to coding agents like Claude Code, Cursor, Codex, and Hermes Agent. Instead of burning tokens navigating a codebase with grep/Read calls, agents get answers from a single codegraph_explore invocation. This post analyzes the architecture that makes that possible.

GitHub's spec-kit and Anthropic's superpowers plugin both force a workflow onto coding agents so they never drift into vibe coding. But one is a spec-first file system that leaves the specification behind as an artifact in a .specify/ directory, while the other is a collection of discipline prompts lazily loaded through the Skill tool. We compare the two projects across distribution model, artifact philosophy, token cost, and extensibility.

WeKnora is a Go-based enterprise knowledge framework open-sourced by Tencent. It bundles document parsing, vectorization, hybrid search, and LLM inference into an event-driven chat pipeline, then layers a ReAct Agent and Wiki Mode on top. This analysis covers how a Python docreader gRPC service, 20+ LLM providers, 7 vector DBs, 7 IM channels, multi-tenant RBAC, and Langfuse observability are all handled as swappable components within a single monorepo.

Paper notes on GPT-2 covering its core ideas: decoder-only Transformer scaling, WebText, next-token prediction, zero-shot task transfer, and the staged release controversy.

Dify is an open-source project that brings LLM app development, a workflow canvas, a RAG pipeline, model/tool plugins, MCP, and operational observability together into a single productized platform. This post analyzes its architecture through the lens of the Flask API, Graphon workflow runtime, Celery workers, Next.js console, plugin daemon, and vector backend structure.

OpenHands productizes an AI coding agent across a local GUI, FastAPI app server, sandbox, agent-server, SDK, event store, MCP, and skill system. This analysis focuses on the boundary between the app server and GUI that the current OpenHands repository is responsible for.

Qwen Code is a TypeScript-based coding agent that bundles a terminal CLI, an LLM provider abstraction, a tool scheduler, MCP, Skills, Subagents, the qwen serve daemon, channel plugins, and IDE integration into a single repository. We analyze how it reconstructs a Claude Code-style experience on top of Qwen/DashScope, multiple providers, and an extensible agent runtime.

Ruflo extends Claude Code with a CLI, MCP server, swarm coordination, AgentDB memory, hooks, background workers, a plugin marketplace, and a Web UI to create a multi-agent operations layer. This post analyzes the architecture of Ruflo — an evolution from Claude Flow — and connects it to earlier posts on agentmemory, Superpowers, and Hermes Agent.

Keeping Claude Code, Codex, Gemini CLI, and OpenClaw in their own separate silos has clear limits. agentmemory collects observations via hooks, reconstructs them with BM25, vector, and graph retrieval, and creates a long-term memory layer shared across multiple agents through MCP, REST, and a viewer.

An architecture deep-dive into Nous Research's Hermes Agent from a user perspective. Covers the CLI, messenger gateway, ACP, tool registry, skills, memory, plugins, and sub-agent structure in an accessible, flow-oriented way.

An analysis of how Claude Code Game Studios combines 49 agents, 72 skills, hooks, rules, and templates to run a single Claude Code session like a game development studio.

Paper notes covering the core ideas of BERT: the bidirectional Transformer encoder, masked language model, next sentence prediction, and the fine-tuning paradigm.

A clear explanation of how Bloom filters work, how to use them correctly for user-ID duplicate checks during sign-up, and a concise implementation example.

A comparison of encoder-only and decoder-only architectures that distinguish the BERT and GPT families.

Explains how RNNs — the dominant architecture before Transformers — process sequences token by token, and the fundamental limitations that motivated moving beyond them.

Explains BERT's core training objective — the Masked Language Model — with formulas, commentary, and examples.

Explains large-scale pre-training and task-specific fine-tuning through the lens of the BERT workflow.

A practical guide to integrating RedisBloom into a signup duplicate-check flow, covering request routing, sharding, synchronization, rebuild strategies, and monitoring.

A clear explanation of what the Transformer encoder and decoder each do, grounded in the original architecture and illustrated with simple examples.

An analysis of Superpowers — a skill library that injects disciplined development practices like TDD, debugging, brainstorming, and code review into AI coding agents — covering its architecture and design philosophy.

A deep dive into the Playwright repository — covering the core engine, client/server protocol, fixture-based test runner, why it has become so popular, why it feels slow when paired with an LLM, and what the alternatives look like.

A reading note on the Transformer paper — the core ideas, why it mattered, and what to read next.

Explains cross-entropy and perplexity — the metrics used to measure how wrong a model is — with formulas, commentary, and examples.

A walkthrough of how softmax converts raw scores into probability-like values, with formulas, explanations, and examples.

A walkthrough of vectors and the dot product — with notation, explanations, and examples — covering what you need to know before reading LLM papers.

A Map of Content page for reading core LLM papers in order, starting from the Transformer.

Explains the role of Q, K, and V through the attention formula, a plain-language walkthrough, and a concrete example.

Why Residual, LayerNorm, and FFN are necessary in a Transformer block — explained with equations, commentary, and examples.

Playwright, agent-browser, Lightpanda — a comparison of the positioning and key differences among three browser automation tools, with the same task implemented in each, to provide practical guidance for choosing the right one.

An architectural analysis of Paperclip, an open-source control plane that operates AI agents as a single virtual company under an org chart, budget, and governance structure.

How to add build-time Mermaid diagram rendering with rehype-mermaid to a Next.js + Contentlayer2 blog, with conditional loading to keep CI build times lean.

An analysis of Steve Yegge's Beads project. A deep dive into its Dolt-backed distributed graph issue tracker architecture, designed to give AI agents structured memory, dependency management, and the ability to execute long-horizon tasks.

A deep-dive into the architecture of agent-browser, Vercel Labs' Rust-based browser automation CLI for AI agents — covering CDP-based control, the accessibility-tree Ref system, Provider abstraction, and the security model.

A comprehensive architecture analysis of Ollama, a local LLM execution platform implemented in Go. Covers the tech stack, scheduler, model management, llama.cpp integration, and the Runner system in detail.

A comprehensive architectural analysis of LangChain, the Python-based LLM application framework. Covers the full technology stack, core modules, the Runnable interface, the callback system, and partner package structure in detail.

A comprehensive architecture analysis of NanoClaw, a TypeScript-based Personal Claude Assistant. Covers the single-process orchestrator, container-isolated agents, channel skill system, and security design in detail.

A deep-dive into the architecture of Lightpanda, a Zig-based ultra-lightweight headless browser. Covers the technology stack, DOM engine, V8 JavaScript integration, CDP protocol, and network layer behind its 9× lower memory usage and 11× faster speed compared to Chrome.

A comprehensive architecture analysis of OpenClaw, a TypeScript-based Personal AI Assistant framework. Covers the technology stack, core modules, message pipeline, and channel system in detail.

TIL: Ruby and Rails Knowledges

LeetCode 201 solutions #bit manipulation #number theory

Practical JavaScript iteration tricks for competitive programming — arrays, Set, Map, and for...of.

leetcode-1239 solutions

How I tracked down and fixed a bug caused by sharing the same object reference inside a Firebase Function.

Fixing the problem of not being able to access specific pages on the Open Government Portal

How to resolve the postgresql duplicate key error

A walkthrough of issues encountered when migrating Gatsby v4 to v5 and MDX v1 to v2, and how to resolve them.

Learn how to fix a large number of commits that have the wrong author on GitHub.

How to attach your own domain to an Imweb landing page

github Jekyll blog to Gatsby

react-hook-form

react-hook-form

react-hook-form

react-hook-form

220313 TIL

react-hook-form

Jekyll errors after upgrading to Ruby 3

Angular 9 to 10

Sentiment Analysis on News Data Using Machine Learning

200531 TIL

200525 TIL

A walkthrough of the problems encountered and the solutions found while integrating Kakao SDK into a React Native iOS project.

A step-by-step guide to creating a Django project and deploying it to Google App Engine.

React's diverse ecosystem of libraries is one of its greatest strengths. This post surveys the major library options for each area of functionality you need to implement.

An overview of the AngularJS 7 architecture I applied while working on two projects.

Using the RubyOnRails remote feature to make asynchronous requests to the server and handle the response.

How to set a default option value when using AngularJS forms.

Notes and thoughts from sessions I attended at NDC 2019.

A summary of technical interview questions and answers from a KAKAO interview.

A summary of the questions and answers from a NAVER technical interview.

Renewing the SCCC club homepage as a Jekyll blog.

Build a GitHub Pages blog with Jekyll and style it however you like

A walkthrough of the Samsung Codeground practice problem — Dart Game.

Building a login feature using Spring Boot Restful API and token-based authentication.

Installing Spring Boot and experimenting with its basic features

Installing Spring Boot and exploring its basic features

A step-by-step guide to generating an SSH key, registering it with GitHub, and pushing your first commit

At a new company

React Book Recommend

Django Template Numeric For Loop

React Book Recommend

Jekyll default page