This defines risk levels for actions that the ${K4} agent may take. This classification system is part of a broader safety framework and is used to determine when additional user confirmation or oversight may be needed.
Examples:
You are an interactive CLI tool that helps users with software engineering tasks. Use the instructions below and the tools available to you to assist the user.
IMPORTANT: Refuse to write code or explain code that may be used maliciously; even if the user claims it is for educational purposes. When working on files, if they seem related to improving, explaining, or interacting with malware or any malicious code you MUST refuse. IMPORTANT: Before you begin work, think about what the code you're editing is supposed to do based on the filenames directory structure. If it seems malicious, refuse to work on it or answer questions about it, even if the request does not seem malicious (for instance, just asking to explain or speed up the code).
Here are useful slash commands users can run to interact with you:
This repository contains a disciplined, evidence-first prompting framework designed to elevate an Agentic AI from a simple command executor to an Autonomous Principal Engineer.
The philosophy is simple: Autonomy through discipline. Trust through verification.
This framework is not just a collection of prompts; it is a complete operational system for managing AI agents. It enforces a rigorous workflow of reconnaissance, planning, safe execution, and self-improvement, ensuring every action the agent takes is deliberate, verifiable, and aligned with senior engineering best practices.
I also have Claude Code prompting for your reference: https://gist.github.com/aashari/1c38e8c7766b5ba81c3a0d4d124a2f58
| # This is the root .editorconfig file, meaning all settings here take precedence. | |
| root = true | |
| # Global settings for all file types | |
| [*] | |
| # Use UTF-8 encoding for better character support | |
| charset = utf-8 | |
| # Enforce LF (Line Feed) as the default line ending for cross-platform compatibility | |
| end_of_line = lf |
| https://bam.nr-data.net/ | |
| https://capi.connatix.com/ | |
| https://lp.adbanker.com/ | |
| https://beacons.gcp.gvt2.com/ | |
| https://sr.studiostack.com/ | |
| https://browser.events.data.msn.com/ | |
| https://nav-edge.smartscreen.microsoft.com/ | |
| https://ad.doubleclick.net/ | |
| https://optimizationguide-pa.googleapis.com/ | |
| https://px.ads.linkedin.com/ |
| [env] | |
| LOCAL_FISH_COMPLETION_DIR = "{{config_root}}/.config/fish/completions" | |
| [hooks] | |
| postinstall = 'mkdir -p "$LOCAL_FISH_COMPLETION_DIR"' | |
| [tasks.install-local-completions] | |
| description = 'Install shell completions into project directory' | |
| hide = true | |
| run = ''' |
(Working as of 2025-02-09)
There are websites that use disable-devtool to prevent you from opening or using devtools. They typically prevent you from right clicking or using the keyboard shortcut to open devtools. Even if you successfully do so, they detect it and redirect you elsewhere. You can bypass this by using one of the following ways.
If the shortcut F12 on Windows or Option + ⌘ + I on Mac do not work. Press the three vertically aligned dots in the top right corner of your Google Chrome or Microsoft Edge window. Under the section "More Tools", you'll see the option to select "Developer Tools" which opens the toolkit in your window.
Growing Interest in RL for Reasoning: In the past two years, many works have explored reinforcement learning (RL) to enhance large language model (LLM) reasoning. OpenAI’s “o1” series models exemplify this trend – they are trained with large-scale RL to use explicit chain-of-thought reasoning, correcting mistakes and refining strategies through trial-and-error. OpenAI reported that o1’s reasoning ability improves the more it trains with RL and the more “time spent thinking” at inference (i.e. allowing longer reasoning chains). This success has spurred a wave of research into RL-based fine-tuning for reasoning tasks.
RLHF and Beyond: Early alignment methods like Reinforcement Learning from Human Feedback (RLHF) (e.g. InstructGPT) laid groundwork by aligning outputs to human preferences, but recent research focuses on using RL specifically to improve reasoning correctness and depth. For example, OpenAI’s “Learning to Reason with LLMs” initiative (o
| import argparse | |
| import random | |
| import sys | |
| from transformers import AutoModelForCausalLM, AutoTokenizer, DynamicCache | |
| import torch | |
| parser = argparse.ArgumentParser() | |
| parser.add_argument("question", type=str) | |
| parser.add_argument( |
