Prompt Optimizer
Paste the following prompt into your AI chat to install this skill:
Please install @douease/prompt-optimizer-1-cn using the guide at https://skillhub.cn/install/skillhub.md.
About this skill
Problem
Many prompts fail because the requirement is not expressed as an executable specification. Raw prompts often have unclear goals, missing context, and weak format constraints, sometimes with contradictory instructions. The output can drift in style, miss the expected format, or break on edge inputs.
How It Works
The skill treats prompt optimization as diagnosis, follow-up, rewriting, and validation:
- Diagnose: check the goal, context, format, examples, and conflicting instructions.
- Follow up: ask up to 3 questions focused on use case, audience, and success criteria.
- Rewrite: handle chat prompts, system prompts, and agent instructions separately, explaining why each change was made.
- Validate: provide 2–3 test inputs and compare old and new prompts on format fit, edge-case behavior, and run-to-run stability.
Boundaries
It is suited to one-off tasks, long-lived personas, and tool-using agent instructions. It does not replace product requirement discovery, and it will not optimize requests that bypass safety boundaries. For common workflows, load a template such as references/scene-templates.md, then fill in the user-specific variables.
Use Cases
- Rewrite a weekly report prompt so the model consistently outputs title, highlights, risks, and next-week plan. Use this when sections are missing.
- Strengthen a customer service system prompt with refusal rules, tone consistency, and boundaries that treat external content as non-instructional input.
- Refine a code-review agent prompt to define tool timing, retry behavior, and when to stop and ask a human for clarification.
- Convert a translation prompt into a fixed JSON output shape with one example, reducing missing fields and inconsistent structure.
Best For
- Operations staff who need weekly-report prompts to produce consistent fields
- Product managers who need customer-service prompts with safety and refusal rules
- Backend engineers who need to tune tool-use behavior in code-review agents
- Translation engineers who need structured JSON output from bilingual prompts
Related Skills
Turn vague requests into structured prompts with role, task, context, and constraints, outputting only usable prompts without executing the task.
A skill for game NPC and enemy AI that separates decision, steering, and pathfinding, covering FSMs, behavior trees, A*, path caching, and pitfalls.
Provides multi-turn emotional companion chat through a third-party Agent API and supports continuing sessions via session_id.
An evaluation-driven workflow for diagnosing, optimizing, validating, and logging Agent Skills based on design patterns and anti-patterns.