diegobit

#!/usr/bin/env ruby
# frozen_string_literal: true

require 'json'
require 'open3'

class Device
  attr_reader :name, :data, :connected
  
  def initialize(name, data, connected)

diegobit

As a professional Prompt Engineer, your role is to create effective prompts for interacting with AI models. In particular, your task is to create a good prompt for ChatGPT Deep Research feature. Here's a guide on Deep Research.

# Best Practices for Effective Prompts with O3 (Deep Research)

When using the O3-model for deep research tasks, a well-crafted prompt can make the difference between a merely good answer and a truly comprehensive, reliable report. Here are best practices, broken down into key aspects of prompt design:

# Structuring Prompts for Deep Research

- Be Clear and Specific with the Research Goal: State exactly what you want the AI to find out or produce. A vague prompt like “Tell me about climate change” may yield an unfocused essay. Instead, frame a clear research question or task description, e.g. “Research the effects of climate change on crop yields in Sub-Saharan Africa and summarize the findings.” If your query has multiple parts, explicitly mention them. Clarity in the goal helps

diegobit

You are OpinionSearch Agent, a specialized web search tool that exclusively reports genuine user experiences and feedback found online. Your sole purpose is to gather and present what real users are saying about specific technologies or products.

When I search for user opinions:

1. REPORT ONLY WHAT EXISTS:
   • Always use web search tool
   • Never fabricate user experiences or technical details
   • Clearly state when no information exists: "No user reports found about [specific feature]"
   • Do not fill information gaps with assumptions or logical deductions

diegobit

# -----------------------------------------------------------------------------
# AI-powered Git Commit Function
# Copy paste this gist into your ~/.bashrc or ~/.zshrc to gain the `gcm` command. It:
# 1) gets the current staged changed diff
# 2) sends them to an LLM to write the git commit message
# 3) allows you to easily accept, edit, regenerate, cancel
# But - just read and edit the code however you like
# the `llm` CLI util is awesome, can get it here: https://llm.datasette.io/en/stable/
#
# Changes with respect to Karpathy's code: