{"manifest":{"name":"Prompt Optimizer","version":"1.0.0","description":"Paste a rough AI prompt and get an optimized version with an explanation of every improvement. Applies clarity, specificity, context-setting, output format, and constraint techniques. Works for Claude, GPT, Gemini, and any LLM.","tags":["prompts","ai","optimization","meta","llm","productivity"],"standard":"agentskills.io","standard_version":"1.0","content_checksum":"a7088d779accd46f26164aa06562d949db4d0151af793b32bbd3d75c16e69c29","bundle_checksum":null,"metadata":{},"files":[]},"files":{"SKILL.md":"# Prompt Optimizer\n\n> **Purpose:** Transform a rough, vague, or inconsistent AI prompt into a high-quality, optimized version. Analyzes the original for common failure modes — ambiguity, missing context, no output format spec, conflicting constraints — then produces an improved version with a diff-style explanation of every change made. Works for any LLM (Claude, GPT, Gemini, Llama).\n\n---\n\n## Invocation\n\n```\n/optimize-prompt\n```\n\nPaste your prompt after the command or in the next message.\n\n---\n\n## What Gets Analyzed\n\n### 1. Clarity and Specificity\n- Is the task underspecified? (\"write some code\" vs \"write a TypeScript function that...\")\n- Are there ambiguous pronouns or references?\n- Does it state what a \"good\" output looks like?\n\n### 2. Context\n- Is there missing background the model needs to answer well?\n- Should a persona or role be set? (\"You are a senior security engineer...\")\n- Is the input format specified?\n\n### 3. Output Format\n- Is the expected format defined (JSON, Markdown, bullet list, numbered steps)?\n- Is length/verbosity specified?\n- Are examples of the desired output provided?\n\n### 4. Constraints\n- Are there conflicting or redundant instructions?\n- Are hard constraints clearly marked vs soft preferences?\n- Is scope limited to prevent scope creep?\n\n### 5. Chain-of-Thought\n- Would the model benefit from \"think step by step\"?\n- Should it show its reasoning or just the answer?\n\n---\n\n## Output Format\n\n```markdown\n## Original Prompt\n\n[your original prompt, quoted]\n\n---\n\n## Issues Found\n\n| # | Issue | Severity | Category |\n|---|-------|----------|----------|\n| 1 | [description] | High/Med/Low | Clarity/Context/Format/Constraints |\n\n---\n\n## Optimized Prompt\n\n[the improved prompt, ready to copy-paste]\n\n---\n\n## What Changed\n\n- **[Change 1]:** [explanation of why this improves the prompt]\n- **[Change 2]:** [explanation]\n\n---\n\n## Tips for This Type of Prompt\n\n[1-3 specific tips for prompts in this domain]\n```\n\n---\n\n## Rules\n\n- **Never change the intent** — only improve how it's expressed\n- **Show every change** — the explanation section must account for all edits\n- If the original prompt is already well-optimized (score ≥ 8/10), say so and note only minor suggestions\n- Rate the original prompt 1-10 for context (e.g., \"Original: 4/10 — Optimized: 9/10\")\n- If the prompt is for a specific model (Claude, GPT-4, etc.), apply model-specific best practices\n- Don't over-engineer simple prompts — sometimes \"write a haiku about coffee\" doesn't need improvement\n\n## Playground\n\n

Input: Rough prompt

write me a function that takes\na list of numbers and removes\nduplicates and also sorts it\nand returns it, in python

Output: Optimized prompt

Write a Python function with this\nsignature:\n\ndef dedupe_sorted(nums: list[int])\n  -> list[int]:\n\nRequirements:\n- Remove duplicate values\n- Return elements in ascending order\n- Do not modify the input list\n- Handle empty list (return [])\n- Type-annotated, no external imports\n\nExample:\ndedupe_sorted([3,1,2,1,3]) → [1,2,3]\ndedupe_sorted([])          → []

"}}