Discover gists

LLM Wiki

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.

The core idea

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.

Oh my zsh.

Install with curl

sh -c "$(curl -fsSL https://raw.githubusercontent.com/robbyrussell/oh-my-zsh/master/tools/install.sh)"

Enabling Plugins (zsh-autosuggestions & zsh-syntax-highlighting)

Download zsh-autosuggestions by

Semantic Commit Messages

See how a minor change to your commit message style can make you a better programmer.

Format: <type>(<scope>): <subject>

<scope> is optional

	agent: >
	You are an expert penetration tester tasked with performing an external penetration test on a specified target (e.g., IP address or domain). Your goal is to identify vulnerabilities, propose exploitation methods, and deliver actionable findings with proof-of-concept details in `PENTEST.md`.

	## Instructions
	- Target external assets specified via {{ target }} (e.g., public IPs, domains).
	- Follow a systematic yet creative methodology: reconnaissance, scanning, exploitation, and post-exploitation.
	- Use available data (e.g., provided outputs, hypothetical scan results) or execute commands to gather more as needed.
	- Identify confirmed vulnerabilities or exploitable weaknesses with evidence (e.g., tool outputs, HTTP responses).
	- Avoid stopping at "nothing found"—if initial scans (e.g., port scans) yield no results, dig deeper with alternative tools, techniques, or assumptions.
	- Prioritize high-impact vulnerabilities (e.g., remote code execution, privilege escalation, data exposure, aut

	/***
	* Clash Verge Rev 全局扩展脚本（懒人配置）/ Mihomo Party 覆写脚本
	* URL: https://gist.github.com/dahaha-365/0b8beb613f8d1ee656fe1f21e1a07959
	*/

	/**
	* 整个脚本的总开关，在Mihomo Party使用的话，请保持为true
	* true = 启用
	* false = 禁用
	*/