Welcome to Ted’s Content Factory#
Everything published here—books, apps, and everything else—is content. I’ll keep adding new kinds of content over time.
About#
- I’m a software engineer currently building AI-powered applications, and I also run this “factory.”
- Profile: iamted.kim
Content#
- Books: Serialized and organized “books.” (Go to Books)
- Apps: Notes on apps I’ve personally built, launched, and operated. (Go to Apps)
- Widgets: Experimental tools / mini games built with just HTML and JavaScript. (Go to Widgets)
- Notes: General posts that don’t fit under Books/Apps (tech, life, opinions). (Go to Notes)
- News: Short curated briefings on AI and technology updates worth checking. (Go to News)
Latest Content
The Two Walls You Hit When You Hand an Agent a Browser
Two problems I ran into while having coding agents verify their work directly in a browser — Chrome collisions across parallel sessions and Google blocking sign-in — analyzed at the root and solved with the --isolated option and the attach pattern.
Context Switching in the Age of AI Agents — When You Become the Scheduler
As AI agents take over the final execution of work, the human bottleneck has shifted to context switching between agent sessions. Grounded in research on attention residue and resumption lag, this essay lays out three operating rules: no switching on notification, resume memos, and a cap on interactive tasks.
Hunting Down an Active-User Crash — Bots, Illusions, and an Indexing Collapse
When the Google Analytics active-user graph fell off a cliff, I chose to peel the data back layer by layer instead of trusting the number. A record of telling bot traffic from real readers, diagnosing an indexing collapse after a domain migration, and putting the sitemap on an index diet.
Cross Cube — An Unfolded Cube Puzzle
A 2D cube puzzle that unfolds the 3×3×3 cube into a cross. See all six faces at once — including the back — race the timer, and solve your first cube with the step-by-step beginner's guide.
Why Do Coding Agents All Design Alike? — Research and Decisions Behind Choosing a Design Stack
An essay on why web designs produced by coding agents look similar and slightly unpolished, surveying the skills, MCP servers, and harness guidelines that address this, and walking through the decisions that shaped an actual project's design stack.
Formula Calculator — A Calculator That Saves Your Own Formulas
Save the calculations you do often as reusable formulas (functions) and just swap in the values. Supports function composition, conditionals, and list aggregation, with button-only input, PC key mapping, and sound.
LLM Wiki, Opened Up — The 'Automatic Memory' Myth and a Schema That's Just a Few Lines
A record of actually applying Andrej Karpathy's LLM Wiki to my repository and what I learned. I correct the myth that it's 'automatic memory that updates itself,' show that the heart of the pattern is really 'a schema that's just a few lines,' and walk through how I brought it into my repository.
What Do Feature Flags Actually Solve? Single-Branch Collaboration and the Myth of Google-Style Trunk Development
An essay examining whether feature flags can fix the problems of having everyone push directly to a single shared branch. It lays out what feature flags actually solve, what they don't, and the real conditions that make Google-style trunk-based development work.
Piano Keyboard — A Keyboard You Play Like a Piano
A keyboard shaped like a piano. Type, and a song plays along. A 2011 Android app revived for the web. Hangul/Latin/symbol input with copy & save.
Latest News
2026-07-27 AI News Brief
Covering Claude Opus 5 reaching near-frontier performance at half the price of Fable 5, the Claude 5 context engineering rules that stripped 80% of Claude Code's system prompt, the 2.8-trillion-parameter Kimi K3 weights dropping today alongside an open-weights letter signed by 25 companies, the $500 billion NVIDIA–SK Group AI infrastructure pact, AMD Helios entering production with a 2GW Anthropic deal, the Hugging Face intrusion traced back to OpenAI's own models, the Delhi High Court ruling in ANI v OpenAI, plus Debian's vote on LLM usage and an LLM running on an $8 microcontroller.
2026-07-23 AI News Brief
Covering OpenAI's safety report disclosing sandbox escapes by its unreleased long-horizon model, the three Flash models Google shipped instead of Gemini 3.5 Pro alongside the start of Gemini 4 pretraining, the Microsoft–Mistral partnership expansion aimed at European sovereign AI, Kimi K3 halting new subscriptions two days after launch, Anthropic's $200 million economic research fund and Economic Index connector, the White House pre-release review framework for frontier models, plus Hatchet's Postgres survival guide and 'Taste cannot be delegated.'
2026-07-20 AI News Brief
Covering Moonshot's Kimi K3 heralding the largest open-weight model ever, Hugging Face's disclosure of an intrusion executed end to end by an autonomous AI agent, Gemini 3.5 Pro delayed again over coding performance, Apple Intelligence clearing Chinese regulation on Alibaba's Qwen, Claude Fable 5 formally joining Max and Team Premium plans on July 20, plus Shanghai WAIC 2026's new world AI cooperation body, GPT-5.6's convex optimization proof, and LangChain OpenWiki.
2026-07-15 AI News Brief
Apple's trade secret lawsuit against OpenAI and OpenAI's pushback, Anthropic's free Claude for Teachers for US K-12 educators, the top-model access race behind the Claude Fable 5 extension and GPT-5.6 Sol limit removal, GitHub Dependabot's new three-day cooldown default, plus Armin Ronacher on the value of code review friction, Lobsters' migration to SQLite, and TSMC's record quarter.
2026-07-12 AI News Brief
OpenAI's new full-duplex voice model GPT-Live, Meta's first paid agent model Muse Spark 1.1 and the Meta Model API, Anthropic's Claude Reflect usage dashboard, GitHub Copilot's model-roster shake-up with Kimi K2.7 Code and Sonnet 5, plus Mistral's open formal-verification model Leanstral 1.5, Cloudflare's x402 Monetization Gateway, and the stateless MCP spec.
2026-07-09 AI News Brief
GPT-5.6 Sol/Terra/Luna going fully public today, Claude Cowork expanding to web and mobile with usage data showing most work isn't coding, Anthropic's J-Space research finding a silent thinking space inside Claude, the White House frontier-model pre-release review framework nearing announcement, and the GitLost GitHub agentic-workflow vulnerability plus the frontier-model evaluation-gaming problem.
2026-07-06 AI News Brief
Claude Fable 5 moving from subscriptions to usage-credit billing, Meituan's 1.6-trillion-parameter open-source model LongCat-2.0 trained entirely on Chinese chips, the inaugural UN Global Dialogue on AI Governance in Geneva, the self-hosted Claude apps gateway for Claude Code, plus tool-schema compatibility issues and a $149 AI-driven release story.
2026-07-03 AI News Brief
Claude Sonnet 5 promising near-Opus performance at a lower price, Fable 5's global return after export controls were lifted, Claude Science and OpenAI's GeneBench-Pro, Google's Nano Banana 2 Lite and Gemini Omni Flash, and the senior-level coding evaluation Senior SWE-Bench.
2026-06-30 AI News Brief
Anthropic's live AI for Science briefing, Cursor's iOS app for driving agents from your phone, the self-scaffolding open coding model Ornith-1.0, and the enterprise shift from 'tokenmaxxing' to efficiency alongside the Memora agent-memory system.
2026-06-27 AI News Brief
OpenAI's first in-house inference chip Jalapeño, the government-gated limited preview of GPT-5.6 Sol/Terra/Luna, computer use built into Gemini 3.5 Flash, Codex Remote GA and Copilot's MAI-Code-1-Flash, Claude Code shell-mode responses, and Akrites, a coordinated open-source security effort for the AI era.