Today's Prompt Engineering: Fastest-Growing Projects — June 11, 2026
This week, the Prompt Engineering space on GitHub continues to heat up with a variety of innovative tools and frameworks designed to enhance the capabilities of AI-driven agents across different domains, such as coding, writing, and creative content generation. Among these projects, "female-portrait-director" stands out for its unique approach to directing detailed AI female portrait prompts.
The "female-portrait-director" repository by liyue-aigc offers a modular Codex Skill designed to direct and expand detailed AI-generated female portraits. With a growth score of 41.04 and over 640 stars, this project is gaining traction due to its specialized focus on enhancing the quality and detail of portrait generation.
"gsd-pi" from open-gsd introduces a powerful meta-prompting system aimed at enabling agents to work autonomously for extended periods without losing context or direction. With a growth score of 33.70, this repository has amassed over 583 stars, reflecting its utility in managing complex tasks and maintaining long-term project focus.
"prompt-refine-skill," developed by Li-Bailiang, is an Agent Skill that silently refines prompts for the currently running model during execution. This tool's growth score of 22.75 and 57 stars indicate a growing interest among developers looking to optimize their AI models' performance in real-time.
"harness-starter-kit" from harnessworks offers a prompt-first approach for making repositories safer for AI coding agents, emphasizing safety and security right from the start. With a growth score of 21.34 and 71 stars, this project highlights the importance of integrating robust engineering practices early in development cycles for projects involving AI.
"gentle-coding," created by OttoRenner, is a PoC demonstrating how empathetic framing can mitigate performance issues in modern LLM reasoning frameworks caused by authoritarian prompt engineering. The repository's growth score of 19.70 and 96 stars suggest that developers are increasingly interested in understanding and addressing the psychological impacts on AI agents.
"vibe-check," developed by TexasBedouin, serves as a guide for beginners to transform vague ideas into executable plans using various AI tools like Claude Code and Codex. With a growth score of 18.42 and 29 stars, this tool is gaining popularity among those looking to streamline their project ideation process.
"seedance-prompts-skill," by mantoufan, provides professional Seedance 2.0 AI video script and storyboard prompt generation for turning articles or stories into screenplays with asset prompts and storyboards. Its growth score of 17.64 and 55 stars indicate growing interest in automated content creation for the media industry.
"journal-adapt-writing-skill," by WantongC, enables users to adapt manuscripts according to the writing conventions of specific journals based on their published papers. This project's impressive star count (673) alongside a growth score of 15.95 highlights its value in academic and scientific publishing sectors.
"prompt-cache-skills," developed by OnlyTerp, provides drop-in prompt-caching fixes for LLM agent harnesses to improve performance and efficiency. With a growth score of 10.79 and 98 stars, this tool is seen as essential for optimizing the operational aspects of AI agents in real-world applications.
"awesome-goal-prompts," created by majiayu000, compiles source-backed /goal contracts and seed patterns to guide coding agents effectively. Its growth score of 8.84 and 33 stars suggest a growing community interest in structured goal-setting for AI-driven projects.
These tools collectively demonstrate the evolving landscape of prompt engineering, with developers focusing on efficiency, safety, creativity, and psychological well-being in their interactions with AI systems.
The "female-portrait-director" repository by liyue-aigc offers a modular Codex Skill designed to direct and expand detailed AI-generated female portraits. With a growth score of 41.04 and over 640 stars, this project is gaining traction due to its specialized focus on enhancing the quality and detail of portrait generation.
"gsd-pi" from open-gsd introduces a powerful meta-prompting system aimed at enabling agents to work autonomously for extended periods without losing context or direction. With a growth score of 33.70, this repository has amassed over 583 stars, reflecting its utility in managing complex tasks and maintaining long-term project focus.
"prompt-refine-skill," developed by Li-Bailiang, is an Agent Skill that silently refines prompts for the currently running model during execution. This tool's growth score of 22.75 and 57 stars indicate a growing interest among developers looking to optimize their AI models' performance in real-time.
"harness-starter-kit" from harnessworks offers a prompt-first approach for making repositories safer for AI coding agents, emphasizing safety and security right from the start. With a growth score of 21.34 and 71 stars, this project highlights the importance of integrating robust engineering practices early in development cycles for projects involving AI.
"gentle-coding," created by OttoRenner, is a PoC demonstrating how empathetic framing can mitigate performance issues in modern LLM reasoning frameworks caused by authoritarian prompt engineering. The repository's growth score of 19.70 and 96 stars suggest that developers are increasingly interested in understanding and addressing the psychological impacts on AI agents.
"vibe-check," developed by TexasBedouin, serves as a guide for beginners to transform vague ideas into executable plans using various AI tools like Claude Code and Codex. With a growth score of 18.42 and 29 stars, this tool is gaining popularity among those looking to streamline their project ideation process.
"seedance-prompts-skill," by mantoufan, provides professional Seedance 2.0 AI video script and storyboard prompt generation for turning articles or stories into screenplays with asset prompts and storyboards. Its growth score of 17.64 and 55 stars indicate growing interest in automated content creation for the media industry.
"journal-adapt-writing-skill," by WantongC, enables users to adapt manuscripts according to the writing conventions of specific journals based on their published papers. This project's impressive star count (673) alongside a growth score of 15.95 highlights its value in academic and scientific publishing sectors.
"prompt-cache-skills," developed by OnlyTerp, provides drop-in prompt-caching fixes for LLM agent harnesses to improve performance and efficiency. With a growth score of 10.79 and 98 stars, this tool is seen as essential for optimizing the operational aspects of AI agents in real-world applications.
"awesome-goal-prompts," created by majiayu000, compiles source-backed /goal contracts and seed patterns to guide coding agents effectively. Its growth score of 8.84 and 33 stars suggest a growing community interest in structured goal-setting for AI-driven projects.
These tools collectively demonstrate the evolving landscape of prompt engineering, with developers focusing on efficiency, safety, creativity, and psychological well-being in their interactions with AI systems.