18 Prompt Engineering tips and tools (all free)
18 Prompt Engineering resources, picked out of 6 saved Instagram reels and 12 liked tweets, from creators like @aitoolhub.co, @gregisenberg, @TheAIColony. 18 of them are free, freemium or open source.
A 68-page whitepaper authored by Lee Boonstra and published by Google covering modern prompt engineering strategies. It provides detailed guidance on model parameters like temperature and top-p, alongside structured prompting frameworks. It serves as a thorough reference for developers looking to improve LLM accuracy and reliability in production.
A practical framework for structuring prompts and selecting input frames when generating video with Kling AI. It emphasizes concise language, a structured sequence covering subject, setting, action, and camera controls, and prioritizing quality initial keyframes over complex text.
A tested collection of over 50 copy-paste prompts designed for Google's Nano Banana Pro image model. It organizes prompts across creative categories like product photography, UI mockups, infographics, and photo editing to reliably achieve studio-quality outputs.
This prompt template guides reasoning models like DeepSeek-R1 to generate complete system instructions for inbound voice receptionists. It outputs identity guidelines, phased conversational workflows, and realistic multi-turn dialogue scenarios for phone-based booking agents.
The Prompt Generator is a built-in developer tool within Anthropic's Claude Console that transforms simple task instructions into optimized, structured system prompts. It automatically adds proper formatting, role definitions, XML tags, and variable placeholders, eliminating the need to manually craft complex prompt syntax.
awesome-nanobanana-pro is a curated GitHub repository collecting ready-to-use prompts, styles, and workflow guides for Google's Gemini image generation model (Nano Banana Pro / Nano Banana 2). It features structured prompts spanning photorealism, retro aesthetics, e-commerce, and photo editing alongside sample visual outputs. It serves as a practical reference for mastering prompt engineering and achieving specific visual styles with Gemini.
A structured ChatGPT system prompt designed to act as a director, cinematographer, and sound designer. It breaks down raw user ideas or images and expands them into cinematic prompts complete with camera movement, subject action, lighting, and audio details. It is tailored for modern video generators like Kling, Runway, and Luma.
This cheat sheet distills key advice from Google's comprehensive prompt engineering guide into thirteen clear rules. It covers foundational techniques like few-shot prompting and chain-of-thought, parameter tuning, context framing, and multi-turn iteration. It serves as a quick reference for improving AI outputs without reading a full technical book.
A collection of Jupyter and Google Colab notebooks created by Matt Shumer for prompt engineering and model distillation. It automatically generates candidate system prompts, benchmarks them using an ELO rating system or task accuracy, and distills large models like Llama 3.1 405B or Claude Opus down to smaller models like Llama 8B or Claude Haiku.
A paired set of prompts designed to generate a stylized landscape base frame in Flux.1 and animate it into a dolly zoom-out video using Seedance. The setup isolates the landscape against a pure white background to make landing page text overlays and scroll-synced hero animations clean and readable.
This workflow lets you extract your natural speaking tone, slang, and delivery style directly from your existing video content. By importing video URLs into Poppy AI and analyzing them with Claude 3.5 Sonnet, you generate a precise tone profile. This profile can then be used as system instructions to draft scripts and content that match your voice.
A structured prompt framework designed to produce high-end, studio-style commercial videos using Google Veo 3. It organizes scene narrative, camera movement, style, lighting, and environmental transitions into a structured JSON schema. The template includes ready-to-use worked examples tailored to major brands like Nike, Tesla, and Louis Vuitton.
A practical prompting method for generating complex WebGL effects and UI animations using AI coding assistants. It breaks the generation process into an initial layout match followed by iterative animation prompting rather than trying to build both at once. It also includes reusable prompt snippets for interactive particles, grids, and border effects.
This note provides a structured prompt template designed for Nano Banana Pro and Gemini image generation to create photorealistic mirror selfies. It breaks down generation rules across subject, wardrobe, pose, scene, lighting, and camera settings. It is particularly useful for learning how to enforce reflection consistency and prevent typical AI distortion errors.
This note provides a structured JSON template for text-to-image models to avoid smooth, artificial-looking skin and overly posed shots. It organizes portrait attributes across clear axes like micro-expressions, hair texture, skin imperfections, and lighting to create natural raw-photo results.
stop-slop is an open-source Claude skill and prompt reference that detects and removes predictable AI writing habits from prose. It provides clear rules against throat-clearing intros, cliché contrast formulas, false agency, and robotic rhythms to make generated text sound natural. It can be installed as a Claude Code skill, added to Claude Project knowledge, or used directly in system prompts.
This repository compiles verbatim system prompts, agent instructions, and tool definitions extracted from major AI chatbots including Anthropic Claude, OpenAI ChatGPT, Google Gemini, and Cursor. It serves as an extensive public archive of real-world prompt engineering, tool use patterns, and personality guidelines used in commercial AI products.
This note provides a structured JSON prompt designed for image generation models like Gemini Nano Banana. It demonstrates how to enforce specific aesthetics—such as early-2000s point-and-shoot flash photography—by categorizing styling, camera emulation, lighting, and subjects into granular key-value pairs.