Today's Prompt Engineering: Fastest-Growing Projects — June 12, 2026
This week, the Prompt Engineering space on GitHub continues to see significant activity as developers and researchers refine techniques for optimizing interactions with AI models through carefully crafted prompts. The liyue-aigc/female-portrait-director project stands out with a high growth score of 38.77, indicating substantial interest in its modular approach to enhancing the quality of AI-generated female portraits. This repository, which boasts over 600 stars, provides tools for directing and expanding detailed prompts aimed at creating more nuanced and accurate AI-generated images.
The open-gsd/gsd-pi project has garnered considerable attention with a growth score of 32.40, accumulating nearly 600 stars in the process. This powerful system is designed to assist agents in maintaining long-term autonomy by engineering meta-prompts and context management strategies that keep them aligned with overarching goals without losing sight of the big picture.
The Li-Bailiang/prompt-refine-skill repository offers a unique approach to prompt refinement, silently enhancing prompts for AI models as they run. With a growth score of 21.36 and an increasing number of commits, this tool is gaining traction among developers looking to improve model performance through subtle and continuous adjustments.
Harnessworks/harness-starter-kit introduces a novel concept in repository safety by focusing on prompt-first engineering techniques aimed at making repositories more secure for AI coding agents. Its growth score of 20.18 reflects the growing interest in methodologies that prioritize security from the outset, with over 74 stars indicating community support.
OttoRenner/Gentle-Coding presents an intriguing exploration into the psychological impact of authoritarian prompt engineering on modern LLM reasoning frameworks. This small-scale proof-of-concept project is drawing attention for its innovative approach to mitigating cognitive issues through empathetic framing strategies, evidenced by a growth score of 18.50 and nearly 100 stars.
TexasBedouin/vibe-check offers an accessible entry point into prompt engineering for non-technical users, guiding beginners from vague ideas to actionable plans with GitHub basics, clean-code habits, and iterative verification loops. With a growth score of 17.36, this project is attracting interest for its broad applicability across different AI platforms.
The mantoufan/seedance-prompts-skill repository focuses on generating high-quality video scripts and storyboards from articles or stories using advanced AI techniques. Its professional approach, detailed guidelines, and integration with official Volcengine resources are contributing to a growth score of 16.79 and over 70 stars.
OnlyTerp/prompt-cache-skills addresses a common challenge in LLM agent harnesses by providing prompt-caching fixes that can be easily integrated into existing setups. This drop-in solution is gaining popularity among developers looking for quick improvements, as indicated by its growth score of 10.13 and over 100 stars.
jnMetaCode/ai-shortfilm-prompts dives deep into the methodology behind creating compelling AI-generated short films through carefully crafted prompts and Claude Code Skills. With a growth score of 7.90 and more than 140 stars, this repository is becoming an essential resource for those interested in pushing the boundaries of AI storytelling.
Finally, zy-zmc/tianming-skill introduces a modular prompt engineering system specifically designed for collaborative long-form novel writing with AI assistance. Its growth score of 5.83 and over 86 stars suggest that it is gaining traction among creative writers who are experimenting with new ways to utilize AI in their work.
These projects collectively illustrate the diverse applications of Prompt Engineering, from enhancing visual creativity and security practices to improving cognitive frameworks for LLMs and driving narrative innovation in long-form writing.
The open-gsd/gsd-pi project has garnered considerable attention with a growth score of 32.40, accumulating nearly 600 stars in the process. This powerful system is designed to assist agents in maintaining long-term autonomy by engineering meta-prompts and context management strategies that keep them aligned with overarching goals without losing sight of the big picture.
The Li-Bailiang/prompt-refine-skill repository offers a unique approach to prompt refinement, silently enhancing prompts for AI models as they run. With a growth score of 21.36 and an increasing number of commits, this tool is gaining traction among developers looking to improve model performance through subtle and continuous adjustments.
Harnessworks/harness-starter-kit introduces a novel concept in repository safety by focusing on prompt-first engineering techniques aimed at making repositories more secure for AI coding agents. Its growth score of 20.18 reflects the growing interest in methodologies that prioritize security from the outset, with over 74 stars indicating community support.
OttoRenner/Gentle-Coding presents an intriguing exploration into the psychological impact of authoritarian prompt engineering on modern LLM reasoning frameworks. This small-scale proof-of-concept project is drawing attention for its innovative approach to mitigating cognitive issues through empathetic framing strategies, evidenced by a growth score of 18.50 and nearly 100 stars.
TexasBedouin/vibe-check offers an accessible entry point into prompt engineering for non-technical users, guiding beginners from vague ideas to actionable plans with GitHub basics, clean-code habits, and iterative verification loops. With a growth score of 17.36, this project is attracting interest for its broad applicability across different AI platforms.
The mantoufan/seedance-prompts-skill repository focuses on generating high-quality video scripts and storyboards from articles or stories using advanced AI techniques. Its professional approach, detailed guidelines, and integration with official Volcengine resources are contributing to a growth score of 16.79 and over 70 stars.
OnlyTerp/prompt-cache-skills addresses a common challenge in LLM agent harnesses by providing prompt-caching fixes that can be easily integrated into existing setups. This drop-in solution is gaining popularity among developers looking for quick improvements, as indicated by its growth score of 10.13 and over 100 stars.
jnMetaCode/ai-shortfilm-prompts dives deep into the methodology behind creating compelling AI-generated short films through carefully crafted prompts and Claude Code Skills. With a growth score of 7.90 and more than 140 stars, this repository is becoming an essential resource for those interested in pushing the boundaries of AI storytelling.
Finally, zy-zmc/tianming-skill introduces a modular prompt engineering system specifically designed for collaborative long-form novel writing with AI assistance. Its growth score of 5.83 and over 86 stars suggest that it is gaining traction among creative writers who are experimenting with new ways to utilize AI in their work.
These projects collectively illustrate the diverse applications of Prompt Engineering, from enhancing visual creativity and security practices to improving cognitive frameworks for LLMs and driving narrative innovation in long-form writing.