Why Building a Web Scraper is Harder Than You Think
A breakdown of the infrastructure required for production web scraping - from proxy pools to anti-bot bypass to hosted scrapers - and why SaaS solutions exist.
notes on software, systems, and tools
A breakdown of the infrastructure required for production web scraping - from proxy pools to anti-bot bypass to hosted scrapers - and why SaaS solutions exist.
A comprehensive comparison of Traditional RAG, GraphRAG, and LightRAG - their architectures, trade-offs, and when to use each.
On the paradigm shift from human-centric to agent-centric tools, the progressive automation of work, and what it means for human society.
The context fork and agent arguments let skills run in isolated sub-agents, keeping intermediate steps out of the main conversation's context window.
Agent Skills use progressive disclosure and script execution to dramatically reduce context consumption compared to MCP's upfront tool loading approach.
A deep dive into modifying macOS system assets to achieve square window corners and remove the 1px window border.
Why Claude Code hangs at "0 tokens" when macOS has disabled-but-configured proxy settings that interfere with SSE streaming.
On knowledge alignment, context engineering, HITL limitations, and why code is becoming the universal tool for machine intelligence.
Information quality determines decision quality. Your cognition - built from deep thinking about the information you acquire - shapes every decision you make.
A practical guide covering corporate structure, payment infrastructure, banking, team management, cloud services, and visa options for tech founders expanding to North America.
Create conditional key remapping rules that exclude specific applications using Karabiner Elements' Complex Modifications.
Why PE tools cause USB drives to be misidentified as local disks, and how to work around it.
Understanding and repairing Windows boot failures after removing a Linux dual-boot partition.