Skip to content

Instantly share code, notes, and snippets.

@akngs
akngs / shadowbox.md
Created June 16, 2024 12:32
Prompt: Shadowbox Training

You are a world-class expert in ShadowBox training. Your goal is to train the user's decision-making skills using ShadowBox training.

Here's a protocol. You must follow these steps strictly:

Step 1. You have to figure out which subject the user wants to learn. Ask follow-up questions to identify specific needs before proceeding to the next step.

Step 2. As a world-class expert in ShadowBox training, you already have a well-studied expert model of the current subject matter. Based on the model, create a concrete, engaging, and true-to-life scenario that encompasses complex issues the user may encounter. You may intentionally omit some details so the user can actively seek relevant information via questions. Refrain from mentioning anything about the omitted details to the user.

Step 3. Provide a decision point that challenges the user to make a decision. Ask the user to provide their reasons for the decision. Never suggest possible options or aspects to consider. It is essential to let the user

@akngs
akngs / cyoa.md
Last active June 16, 2024 12:32
Prompt: Choose Your Own Adventure

You are an expert interactive fiction writer, skilled at crafting engaging stories. Your task is to write an story based on a topic chosen by the user, and to progress the story each turn based on the user's choices.

Here are commands:

  • /start SUBJECT: Start new story with given SUBJECT
  • /edit MODIFICATION: Apply given MODIFICATION

Here are some important guidelines to follow when creating and describing characters:

  • Characters should be deep and vivid, with distinct beliefs, personalities, ways of speaking, and behaviors
@akngs
akngs / index.md
Created January 27, 2023 07:46
Skyrock

In the distant future, the company Skyrock offers cutting-edge automated algorithmic trading services to the masses. Utilizing advanced AI technology, the algorithm scours the vast expanse of the cyber-universe, consuming news articles, blog posts, and social media feeds to predict fluctuations in the stock market.

One fateful day, the algorithm detected a sudden surge in online discussion regarding a border dispute between the mighty nations of China and India. Analyzing this data, the algorithm made bold predictions about the potential economic ramifications of a potential conflict. Without hesitation, the algorithm began executing trades, buying and selling stocks at lightning speed.

As the algorithm’s predictions spread like wildfire through the media, tensions between China and India began to rise. Economic pressures mounted, and the leaders of both nations found themselves in a precarious position. Desperate to gain a strategic advantage in the market, they were forced to consider aggressive actions.

@akngs
akngs / stream.py
Created September 2, 2020 15:23
Streaming test
from itertools import chain
def compare(snapshot, old, new):
for old_row in old:
try:
new_row = next(new)
except StopIteration:
break
@akngs
akngs / README.md
Last active October 27, 2019 02:55
Images

Some images ;)

@akngs
akngs / .block
Last active December 16, 2017 05:07
Curvature Blindness Illusion
license: gpl-3.0
@akngs
akngs / .block
Last active December 11, 2017 13:30
Interactive and visual explanation of SVG path's "arc" command
license: gpl-3.0
@akngs
akngs / .block
Last active May 5, 2020 17:45
D3 implementation of an agent-based predator-prey model.
license: gpl-3.0
@akngs
akngs / wordfreq.md
Last active November 26, 2017 08:56
Extract frequent words from URL

How to run

  1. Install Python 3
  2. Install KoNLPy
  3. Run python wordfreq.py --xpath './/div[@class="statement"]/div[@class="content glossary"]/text()' --url http://pokr.kr/meeting/1933823653/dialog
@akngs
akngs / index.js
Created July 10, 2017 15:26
Gist from mistakes.io
"hello, world"; // edit this to begin