Today's Prompt Engineering: Fastest-Growing Projects — June 09, 2026
Today's the Prompt Engineering space, there's a notable trend towards more specialized and sophisticated tools designed to enhance AI-driven creativity and productivity. These tools range from refining detailed portrait prompts for female characters to adapting manuscripts for specific journal styles. The repository "liyue-aigc/female-portrait-director" stands out with its modular approach to directing detailed AI-generated portraits, gaining significant attention in recent weeks.
The "female-portrait-director" tool by liyue-aigc offers a modular Codex Skill for guiding and expanding detailed female portrait prompts. Its growth score of 42.35 and over 500 stars reflect its popularity among developers looking to refine AI-generated art specifically tailored towards female portraits.
The "gsd-pi" repository from open-gsd is a powerful meta-prompting system designed to keep agents working autonomously without losing sight of the big picture, achieving a growth score of 36.06 and accumulating nearly 540 stars. This tool's robustness in managing long-term tasks through context engineering makes it particularly appealing for projects requiring continuous AI involvement.
Li-Bailiang’s "prompt-refine-skill" is an Agent Skill that silently refines prompts for the currently running model, aiming to improve the quality of generated outputs without user intervention. With a growth score of 29.88 and 47 stars, it appeals to users seeking subtle improvements in AI-generated content through automated prompt refinement.
The "seedance-prompts-skill" by mantoufan provides professional Seedance 2.0 AI video script and storyboard generation capabilities, turning articles or stories into screenplays with asset prompts and storyboards. Its growth score of 20.72, alongside 40 stars, indicates its growing popularity among creators looking to automate the process of converting written content into visual scripts.
OttoRenner’s "Gentle-Coding" showcases how empathetic framing can mitigate performance anxiety in modern LLM reasoning frameworks through a small-scale proof of concept. With a growth score of 18.42 and 92 stars, it addresses an important aspect of AI interaction that enhances user experience by reducing cognitive overload.
WantongC’s "journal-adapt-writing-skill" learns the writing conventions of specific journals from their published papers to help users adapt their manuscripts accordingly. This tool has seen a growth score of 16.78 and garnered over 650 stars, highlighting its utility in academic and professional writing contexts where conformity with journal standards is critical.
The "prompt-cache-skills" repository by OnlyTerp offers drop-in fixes for prompt caching in LLM agent harnesses, aiming to improve efficiency and performance. With a growth score of 12.62 and nearly 100 stars, it addresses the technical needs of developers working with AI coding agents.
"Majiayu000’s awesome-goal-prompts" compiles a collection of /goal contracts for coding agents along with seed patterns to guide their development. Its relatively lower growth score of 9.50 but steady number of commits indicates its ongoing relevance as a resource for prompt engineering in coding tasks.
The "ai-shortfilm-prompts" repository by jnMetaCode details the methodology and prompts used behind an AI-generated short film, demonstrating the potential of AI in creative storytelling. With a growth score of 8.45 and over 120 stars, it showcases how AI can be harnessed to produce compelling narrative content.
Lastly, "zy-zmc’s tianming-skill" is a modular prompt engineering system for collaborative novel writing, gaining traction with a growth score of 6.88 and around 84 stars. This tool emphasizes the creative potential of AI in long-form content creation through structured collaboration.
These tools collectively illustrate the diversity and sophistication of Prompt Engineering efforts this week, each addressing unique challenges within their respective domains while achieving varying levels of community engagement and technical interest.
The "female-portrait-director" tool by liyue-aigc offers a modular Codex Skill for guiding and expanding detailed female portrait prompts. Its growth score of 42.35 and over 500 stars reflect its popularity among developers looking to refine AI-generated art specifically tailored towards female portraits.
The "gsd-pi" repository from open-gsd is a powerful meta-prompting system designed to keep agents working autonomously without losing sight of the big picture, achieving a growth score of 36.06 and accumulating nearly 540 stars. This tool's robustness in managing long-term tasks through context engineering makes it particularly appealing for projects requiring continuous AI involvement.
Li-Bailiang’s "prompt-refine-skill" is an Agent Skill that silently refines prompts for the currently running model, aiming to improve the quality of generated outputs without user intervention. With a growth score of 29.88 and 47 stars, it appeals to users seeking subtle improvements in AI-generated content through automated prompt refinement.
The "seedance-prompts-skill" by mantoufan provides professional Seedance 2.0 AI video script and storyboard generation capabilities, turning articles or stories into screenplays with asset prompts and storyboards. Its growth score of 20.72, alongside 40 stars, indicates its growing popularity among creators looking to automate the process of converting written content into visual scripts.
OttoRenner’s "Gentle-Coding" showcases how empathetic framing can mitigate performance anxiety in modern LLM reasoning frameworks through a small-scale proof of concept. With a growth score of 18.42 and 92 stars, it addresses an important aspect of AI interaction that enhances user experience by reducing cognitive overload.
WantongC’s "journal-adapt-writing-skill" learns the writing conventions of specific journals from their published papers to help users adapt their manuscripts accordingly. This tool has seen a growth score of 16.78 and garnered over 650 stars, highlighting its utility in academic and professional writing contexts where conformity with journal standards is critical.
The "prompt-cache-skills" repository by OnlyTerp offers drop-in fixes for prompt caching in LLM agent harnesses, aiming to improve efficiency and performance. With a growth score of 12.62 and nearly 100 stars, it addresses the technical needs of developers working with AI coding agents.
"Majiayu000’s awesome-goal-prompts" compiles a collection of /goal contracts for coding agents along with seed patterns to guide their development. Its relatively lower growth score of 9.50 but steady number of commits indicates its ongoing relevance as a resource for prompt engineering in coding tasks.
The "ai-shortfilm-prompts" repository by jnMetaCode details the methodology and prompts used behind an AI-generated short film, demonstrating the potential of AI in creative storytelling. With a growth score of 8.45 and over 120 stars, it showcases how AI can be harnessed to produce compelling narrative content.
Lastly, "zy-zmc’s tianming-skill" is a modular prompt engineering system for collaborative novel writing, gaining traction with a growth score of 6.88 and around 84 stars. This tool emphasizes the creative potential of AI in long-form content creation through structured collaboration.
These tools collectively illustrate the diversity and sophistication of Prompt Engineering efforts this week, each addressing unique challenges within their respective domains while achieving varying levels of community engagement and technical interest.