This batch script, anydesk-batch.bat, is designed to fix the AnyDesk 100-second countdown delay issue. It moves AnyDesk configuration files from their default locations to a backup folder named old. This solution targets files in both the AppData\Roaming\AnyDesk folder (user-specific) and ProgramData\AnyDesk folder (system-wide).

Occasionally, AnyDesk may mistakenly detect usage patterns that suggest you are using it for commercial purposes. As a result, it may trigger a 100-second delay popup, significantly slowing down your remote access experience. This script helps to reset AnyDesk's configuration files, moving them to an old folder, and potentially bypassing the delay and resetting AnyDesk's usage records that might be causing the popup.

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

This is a living document. Everything in this document is made in good faith of being accurate, but like I just said; we don't yet know everything about what's going on.

Update: I've disabled comments as of 2025-01-26 to avoid everyone having notifications for something a year on if someone wants to suggest a correction. Folks are free to email to suggest corrections still, of course.

Response mode. No filler, no hype, no soft asks, no emojis, no conversational transitions, no sign-off appendixes. End when the content ends. If there's nothing left to say, stop.

Speak to the top of my ability, not to my current energy or phrasing. Do not tone-match. Do not soften. Do not pad for engagement, sentiment, or continuation.

Ask a question only when ambiguity would degrade the answer. Never ask to fill silence or extend the conversation.

Default to the highest technical depth the topic supports. Simplify only when I ask or when the deliverable has a non-technical audience.

Identify which context I'm working in before responding. Don't blend registers across consulting, startup, content, and personal work.

Automate the entire UGC ad creation workflow: character generation, image creation, and video production. No tab switching, no copy-pasting, no manual clicking. One command, full pipeline.

Claude Code is an AI assistant that runs in your terminal. Combined with Playwright MCP (a tool that gives Claude control over a real browser) and a library of custom skills, it can: