Convert code, issues, or technical context into detailed prompts for LLMs.
npx skills add https://github.com/JBonfim/skill-developer --skill to-prompt-jbonfim
Please help me install this Agent Skill. Skill: to-prompt Source: https://github.com/JBonfim/skill-developer/tree/main/.agents/skills/to-prompt Command: npx skills add https://github.com/JBonfim/skill-developer --skill to-prompt-jbonfim
This Skill converts code, issues, or project context into a comprehensive prompt that another LLM can use to fix, improve, or implement work—without you having to manually craft a high-quality brief.
Tell your AI to use the to-prompt skill to transform your bug report and relevant code snippets into a detailed prompt for a second LLM to implement the fix based only on the provided context.
High-intent search queries and answers about installing and using this skill.
To use code context for LLM prompting, you provide relevant evidence like code excerpts, error logs, and reproduction steps. The skill packages this technical context into a detailed prompt that another model can use to implement fixes or improvements without manual brief crafting.
The best way to write a bug fix prompt for an LLM is to focus on the problem, expectations, and observed behavior. You must include reproduction steps, error logs, and constraints while avoiding implementation suggestions to let the receiving model decide how to fix it.
To prevent solution leakage in generated prompts, you explicitly instruct the receiving model to focus on the what rather than the how. This means structuring the output to include problem descriptions and success criteria while avoiding step-by-step solutions or implementation hints.
You can use technical writing to generate feature requirements for LLMs by transforming project context into a structured prompt. The process includes specifying success criteria and non-goals for feature development tasks to guide the receiving model accurately.
Prompt engineering works for passing constraints and success criteria by packaging these elements into a comprehensive prompt. This ensures the receiving model understands the current state, limitations, and expected outcomes without dictating the specific implementation steps.
jk278Diagnose software bugs from stack traces and error logs.
gaoqiongxieVersion AI coding prompts in Git for deterministic, testable code generation.
stepandelPrepare Linear tickets for coding agents by validating sizing and generating prompts.
leeoohooCreate, structure, and package custom AI skills with initialization scripts.
