modern-prompting

What problem does it solve?

Modern Prompting & Context Engineering Framework enables structured, repeatable prompting strategies to guide large language models and improve reliability.

Core Features & Use Cases

• OOReDAct: Deep deliberation before action and decision to improve task outcomes.
• Chain-of-Thought (CoT): Explicit reasoning steps to boost accuracy on complex problems.
• Chain of Draft (CoD): Iterative, dense summaries that preserve key details.
• Reasoning and Acting (ReAct): Combine thinking with action to interact with tools and environments.
• Pal (Program-Aided Language): Generate and run executable code for deterministic computations.
• Reflexion: Self-improvement loop to refine approaches after mistakes.
• ToT-lite (Tree of Thoughts): Explore multiple reasoning paths before final conclusions.
• Metacognitive Prompting (MP): Structured metacognition for goal clarification and planning.
• Automated Prompt Optimization (APO): Evolve prompts through autonomous experimentation.
• Cache-Augmented Generation (CAG): Preload and manage context to reduce latency.
• Cognitive Scaffolding Prompting: Structured support levels to guide problem solving.
• Advanced Techniques: Internal Knowledge Synthesis (IKS), Multimodal Synthesis, KSP, and more for cross-domain coherence.

Quick Start

Begin by selecting an prompting strategy (for example, CoT or ReAct) and structure your task using the OOReDAct workflow to guide the model's reasoning and actions.