KleoSr CLI — Adaptive Intelligence for Prompt Engineering

KleoSr PE²-CLI: Adaptive Intelligence for Prompt Engineering

Showcase for Anysphere & Cursor Forum

Welcome to the KleoSr PE²-CLI showcase! This lightweight command-line tool is designed to help developers, prompt engineers, and AI enthusiasts craft high-performance prompts for large language models with minimal effort.

:magnifying_glass_tilted_left: What Is PE²-CLI?

PE²-CLI (Prompt Engineering 2.0) uses adaptive intelligence and automated complexity analysis to transform your raw text into optimized prompts. By breaking down each prompt into Context, Role, Task, Constraints, and Output sections, PE²-CLI ensures consistent, high-quality results across a wide range of AI providers.

:rocket: Key Features

  • Adaptive Complexity Scoring: Automatically assesses prompt difficulty (NOVICE → MASTER) and adjusts iteration counts for optimal refinement.
  • Multi-Provider Support: Seamlessly integrates with OpenAI, Anthropic, Google Gemini, OpenRouter, and more.
  • Interactive Mode: Intuitive arrow-key navigation and clean ASCII art banner for a delightful CLI experience.
  • Direct & File Input: Run ad-hoc prompts or feed text files for batch processing.
  • Customizable Workflows: Fine-tune model selection, iteration count, and output formats with simple flags.

:bullseye: Why Choose PE²-CLI?

PE²-CLI accelerates prompt iteration, saving you time and maximizing the impact of your AI calls. Whether you’re prototyping a new chatbot, analyzing data with AI, or teaching prompt engineering best practices, this tool streamlines your workflow and delivers reliable, reproducible prompts.

:hammer_and_wrench: Getting Started

npm install -g @kleosr/pe2-cli
pe2-cli "Your initial prompt here"

For detailed documentation, visit the GitHub repository and check the full README.


Made with :heart: for the Cursor Community by KleoSr

1 Like

Interesting, I might try this before cursor implements (hopefully) their prompt optimizer. Do you have any examples of input and output? Also your github link is wrong. Is there any chance you could turn this into an mcp to be used in cursor directly with the usage I pay for in cursor? Thanks fro sharing!

can you show examples?

Hey there, @liquefy & @2jfs904judsw20600jik

Yes, absolutely! The MCP server will be configured and updated soon. The input and output examples will be uploaded shortly as well and thanks for catching the GitHub link issue.

As for turning this into an MCP tool for Cursor, that’s definitely something I’m working into.

Best regards,
kleosr.

1 Like