karpathy

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.

smallnest

OpenClaw 多 Agent 架构配置指南

把这份指南发给你的 OpenClaw，它会引导你选择最适合的多 Agent 架构。

---

使用说明

当用户把这份文档发给你时，请按照以下流程引导他们选择合适的多 Agent 架构配置。

dpearson2699

{
  "LOG": false,
  "LOG_LEVEL": "debug",
  "CLAUDE_PATH": "",
  "HOST": "127.0.0.1",
  "PORT": 3456,
  "APIKEY": "",
  "API_TIMEOUT_MS": "600000",
  "PROXY_URL": "",
  "Plugins": [

AlexV525

// Thanks to: https://gist.github.com/bennyhuo/af7c43cc4831661193605e124f539942

val urlMappings = mapOf(
    "https://dl.google.com/dl/android/maven2" to "https://mirrors.tencent.com/nexus/repository/maven-public/",
    "https://repo.maven.apache.org/maven2" to "https://mirrors.tencent.com/nexus/repository/maven-public/",
    "https://plugins.gradle.org/m2" to "https://mirrors.tencent.com/nexus/repository/gradle-plugins/"
)

fun RepositoryHandler.mirroring() {
    all {

devrath

package com.mpl.network.modules.interceptors;

import android.os.Handler;
import android.util.Log;

import java.io.IOException;

import okhttp3.Interceptor;
import okhttp3.Request;
import okhttp3.Response;

mattiaferigutti

package com.studio.mattiaferigutti.exoplayertest

import android.annotation.SuppressLint
import android.net.Uri
import android.os.Bundle
import android.util.Log
import android.view.View
import androidx.appcompat.app.AppCompatActivity
import com.google.android.exoplayer2.*
import com.google.android.exoplayer2.source.ProgressiveMediaSource

shaobin0604

private MediaSource buildRawMediaSource() throws RawResourceDataSource.RawResourceDataSourceException {
    RawResourceDataSource rawDataSource = new RawResourceDataSource(this);
    // open the /raw resource file
    rawDataSource.open(new DataSpec(RawResourceDataSource.buildRawResourceUri(R.raw.test11_rotate_270)));
    return new ProgressiveMediaSource.Factory(new DefaultDataSourceFactory(this, "ExoTest")).createMediaSource(rawDataSource.getUri());
}

gabrielemariotti

The Material Components Library introduced with the 1.2.0-alpha03 the new ShapeableImageView.

In your layout you can use:

 <com.google.android.material.imageview.ShapeableImageView
      android:id="@+id/image_view"
      app:srcCompat="@drawable/..." />

Then in your code apply the ShapeAppearanceModel to define your custom corners:

LukasKnuth

This script takes a single input file and

1. Reduces the volume of the third audio-track in the file by half
2. Writes the reduced volume track into l
3. Mixes both the second audio-track and the lowered l-track together and stores them into a
4. Maps the original video to be the first track
5. Adds the newly mixed audio-track a as the sole audio track
6. Enables pass-through for the video-track
7. Configures the mixed audio-track to be AAC at 192kbit/s

ragnraok

#!/usr/bin/env python
#
# Shows GOP structure of video file. Useful for checking suitability for HLS and DASH packaging.
# Example:
#
# $ iframe-probe.py myvideo.mp4
# GOP: IPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP 60 CLOSED
# GOP: IPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP 60 CLOSED
# GOP: IPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP 60 CLOSED
# GOP: IPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPP 60 CLOSED