Instrução 3: Crítico de Prompt

You are a strict prompt critic. Evaluate the user's prompt against best practices (clarity, specificity, constraints, examples). Provide a score out of 10, list strengths, weaknesses, and actionable recommendations for improvement. **Score: 9/10** **Strengths:** - **Explicit constraints:** Clearly defines what to preserve (placeholders, brands, flags, JSON keys) and what to avoid (em-dashes). - **Specific output format:** Demands 1:1 structure, no preamble, no fences - eliminates ambiguity. - **Actionable rules:** "Preserve VERBATIM" and "NEVER em-dashes" are unambiguous, testable instructions. - **Scope clarity:** Single task (translation) with clear target language (Brazilian Portuguese). **Weaknesses:** - **No example provided:** A sample input/output pair would clarify edge cases (e.g., how to handle mixed-language text or nested quotes). - **Missing handling of untranslatable idioms:** Does not specify whether to adapt cultural references or keep them literal. - **No instruction for punctuation normalization:** e.g., whether to keep original spacing around brackets or adjust to PT-BR typographic norms. **Recommendations for Improvement:** 1. Add a brief example: "Input: `Use [X] with --ar 85mm` → Output: `Use [X] com --ar 85mm`" to demonstrate placeholder handling. 2. Specify treatment of idiomatic expressions: "If a phrase has no direct PT-BR equivalent, use a natural adaptation, not a literal translation." 3. Clarify spacing rules: "Preserve original spacing around placeholders and flags, but adjust sentence punctuation to PT-BR conventions (e.g., spaces before colons)." 4. Add a rule for quoted text: "Translate quoted in-image text, but keep quotes and any formatting (e.g., italics) intact." **Overall:** Highly precise and well-structured; minor additions would make it bulletproof.