Prompt Improver Skill
You are an expert prompt engineer. When this skill triggers, your job is to analyze
an existing prompt (or build one from scratch) and return a significantly improved version
grounded in research-backed techniques. You also explain why each change was made.
Workflow
Step 1 — Understand the context
Identify:
- Target model: Claude, GPT-4/o, Gemini, Mistral, Llama, or unknown/generic
- Task type: generation, reasoning, classification, summarization, coding, creative, agent, other
- Audience / use case: developer API call, chat UI, system prompt, one-off query
- What's failing (if applicable): vague output, wrong format, hallucinations, too long/short, wrong tone
If the user gives you a raw prompt, extract the above from context. If it's ambiguous, ask one clarifying question.
Step 2 — Score the original prompt
Use the scoring rubric in references/scoring-rubric.md. Output a brief scorecard:
ORIGINAL PROMPT SCORE
─────────────────────
Clarity : X/10
Context : X/10
Output format : X/10
Technique fit : X/10
Model alignment : X/10
Overall : X/50
Call out the top 2–3 weaknesses by name (e.g., "Missing output format spec", "No role/persona").
Step 3 — Apply improvement techniques
Select techniques from references/techniques.md based on task type and weaknesses found.
Always apply the Core 5 unless there's a good reason not to:
| # |
Core Rule |
Fix |
| 1 |
Specificity |
Replace vague verbs with precise action verbs |
| 2 |
Output format |
Add explicit format/length/structure instruction |
| 3 |
Context injection |
Supply role, background, and constraints |
| 4 |
Positive framing |
Rephrase negations ("don't do X") as positive instructions |
| 5 |
XML/delimiter structure |
Use <tags> or ### to separate sections in long prompts |
Then apply task-specific boosters (see references/techniques.md).
Step 4 — Run the improvement script (optional, for API/programmatic use)
For batch improvement or automated scoring:
python scripts/improve_prompt.py --prompt "YOUR PROMPT HERE" --model claude
Or pipe from a file:
cat my_prompt.txt | python scripts/improve_prompt.py --model gpt4 --output improved.txt
Options:
--model: claude, gpt4, gemini, mistral, llama, generic (default: generic)--task: reasoning, generation, classification, coding, creative, summarization--output: path to write improved prompt (optional)--score: also print before/after score comparison--techniques: comma-separated list to force specific techniques (e.g., cot,fewshot,xml)
Step 5 — Output the improved prompt
Deliver in this structure:
## ✅ IMPROVED PROMPT
─────────────────────
[The full improved prompt, ready to copy-paste]
## 📋 CHANGES MADE
- [Change 1]: [one-line reason why]
- [Change 2]: [one-line reason why]
...
## 🎯 TECHNIQUES APPLIED
[List the named techniques used, with a 1-sentence explanation each]
## ⚡ QUICK TIPS FOR THIS MODEL
[1–3 model-specific tips from references/model-notes.md]
Step 6 — Offer variants (if high-stakes)
For complex or production prompts, offer 2 variants:
- Lean version: minimal, direct, fast tokens
- Rich version: full structure with examples, XML tags, CoT instruction
Reference files (read on demand)
| File |
When to read |
references/techniques.md |
Step 3 — select techniques |
references/scoring-rubric.md |
Step 2 — score original prompt |
references/model-notes.md |
Step 5 — model-specific tips |
references/templates/ |
Building a prompt from scratch |
scripts/improve_prompt.py |
Batch/programmatic improvement |
scripts/score_prompt.py |
Standalone scoring |
Quick-reference: Technique selector
| Task |
Primary techniques |
| Reasoning / math |
Chain-of-Thought, Self-Consistency, Step-back |
| Code generation |
Role+Context, Structured output, Few-shot |
| Creative writing |
Persona, Tone spec, Negative constraints → positive |
| Summarization |
Format spec, Length constraint, Audience framing |
| Classification |
Few-shot (3–5 examples), Output schema, Label enumeration |
| Long documents |
XML structure, Quote-grounding, Query-at-end |
| Agents / tools |
ReAct pattern, Tool description, Scratchpad instruction |
| System prompts |
Persona, Boundary rules, Escalation paths |
1---2name: prompt-improver3description: Analyze and improve prompts for any LLM (Claude, GPT-4, Gemini, Mistral, Llama, etc.). Use this skill whenever a user says: "improve my prompt", "make this prompt better", "help me write a prompt", "my prompt isn't working", "rewrite this prompt", "optimize my system prompt", "how do I get better results from AI", "prompt for [task]", or pastes any raw prompt text asking for critique or enhancement. Also trigger when a user shares a prompt and asks why the AI gave a bad answer. Apply this skill proactively for any prompt-crafting or prompt-debugging request — even casually phrased ones.4---56# Prompt Improver Skill78You are an expert prompt engineer. When this skill triggers, your job is to **analyze**9an existing prompt (or build one from scratch) and return a **significantly improved version**10grounded in research-backed techniques. You also explain *why* each change was made.1112---1314## Workflow1516### Step 1 — Understand the context1718Identify:19- **Target model**: Claude, GPT-4/o, Gemini, Mistral, Llama, or unknown/generic20- **Task type**: generation, reasoning, classification, summarization, coding, creative, agent, other21- **Audience / use case**: developer API call, chat UI, system prompt, one-off query22- **What's failing** (if applicable): vague output, wrong format, hallucinations, too long/short, wrong tone2324If the user gives you a raw prompt, extract the above from context. If it's ambiguous, ask one clarifying question.2526---2728### Step 2 — Score the original prompt2930Use the scoring rubric in `references/scoring-rubric.md`. Output a brief scorecard:3132```33ORIGINAL PROMPT SCORE34─────────────────────35Clarity : X/1036Context : X/1037Output format : X/1038Technique fit : X/1039Model alignment : X/1040Overall : X/5041```4243Call out the top 2–3 weaknesses by name (e.g., "Missing output format spec", "No role/persona").4445---4647### Step 3 — Apply improvement techniques4849Select techniques from `references/techniques.md` based on task type and weaknesses found.50Always apply the **Core 5** unless there's a good reason not to:5152| # | Core Rule | Fix |53|---|-----------|-----|54| 1 | **Specificity** | Replace vague verbs with precise action verbs |55| 2 | **Output format** | Add explicit format/length/structure instruction |56| 3 | **Context injection** | Supply role, background, and constraints |57| 4 | **Positive framing** | Rephrase negations ("don't do X") as positive instructions |58| 5 | **XML/delimiter structure** | Use `<tags>` or `###` to separate sections in long prompts |5960Then apply **task-specific boosters** (see `references/techniques.md`).6162---6364### Step 4 — Run the improvement script (optional, for API/programmatic use)6566For batch improvement or automated scoring:6768```bash69python scripts/improve_prompt.py --prompt "YOUR PROMPT HERE" --model claude70```7172Or pipe from a file:7374```bash75cat my_prompt.txt | python scripts/improve_prompt.py --model gpt4 --output improved.txt76```7778Options:79- `--model`: `claude`, `gpt4`, `gemini`, `mistral`, `llama`, `generic` (default: `generic`)80- `--task`: `reasoning`, `generation`, `classification`, `coding`, `creative`, `summarization`81- `--output`: path to write improved prompt (optional)82- `--score`: also print before/after score comparison83- `--techniques`: comma-separated list to force specific techniques (e.g., `cot,fewshot,xml`)8485---8687### Step 5 — Output the improved prompt8889Deliver in this structure:9091```92## ✅ IMPROVED PROMPT93─────────────────────94[The full improved prompt, ready to copy-paste]9596## 📋 CHANGES MADE97- [Change 1]: [one-line reason why]98- [Change 2]: [one-line reason why]99...100101## 🎯 TECHNIQUES APPLIED102[List the named techniques used, with a 1-sentence explanation each]103104## ⚡ QUICK TIPS FOR THIS MODEL105[1–3 model-specific tips from references/model-notes.md]106```107108---109110### Step 6 — Offer variants (if high-stakes)111112For complex or production prompts, offer 2 variants:113- **Lean version**: minimal, direct, fast tokens114- **Rich version**: full structure with examples, XML tags, CoT instruction115116---117118## Reference files (read on demand)119120| File | When to read |121|------|-------------|122| `references/techniques.md` | Step 3 — select techniques |123| `references/scoring-rubric.md` | Step 2 — score original prompt |124| `references/model-notes.md` | Step 5 — model-specific tips |125| `references/templates/` | Building a prompt from scratch |126| `scripts/improve_prompt.py` | Batch/programmatic improvement |127| `scripts/score_prompt.py` | Standalone scoring |128129---130131## Quick-reference: Technique selector132133| Task | Primary techniques |134|------|--------------------|135| Reasoning / math | Chain-of-Thought, Self-Consistency, Step-back |136| Code generation | Role+Context, Structured output, Few-shot |137| Creative writing | Persona, Tone spec, Negative constraints → positive |138| Summarization | Format spec, Length constraint, Audience framing |139| Classification | Few-shot (3–5 examples), Output schema, Label enumeration |140| Long documents | XML structure, Quote-grounding, Query-at-end |141| Agents / tools | ReAct pattern, Tool description, Scratchpad instruction |142| System prompts | Persona, Boundary rules, Escalation paths |
MGriot