Перейти к содержимому
TakoAPI
Назад к навыкам

senior-prompt-engineer

от alirezarezvani

Coding Agents & IDEs
1.2kЧисло загрузок по данным clawskills.sh
3Звёзды
0Просмотры
0 лайковВойдите, чтобы поставить лайк навыку
—Ещё нет оценок
Ваша оценка

Кратко

This skill should be used when the user asks to "optimize prompts", "design prompt templates", "evaluate LLM.

Описание

Что делает этот навык

Tools and workflows for analyzing LLM prompts, evaluating RAG pipelines, and validating agentic systems. Includes Python scripts for token counting, clarity scoring, few-shot example management, and agent workflow visualization.

Combines prompt analysis, RAG evaluation, and agent validation into a single workflow rather than requiring separate tools for each concern.

Когда его использовать

  • Cutting token costs in a production GPT-4 prompt
  • Measuring retrieval accuracy in a RAG-based Q&A system
  • Validating agent tool configs before a deployment
  • Building consistent few-shot examples for entity extraction
  • Enforcing structured JSON output from an LLM

Пример сценария

ВВОД

User asks: Optimize this customer support prompt to reduce token usage

АГЕНТ
  1. 1

    Runs prompt_optimizer.py --analyze on the prompt file to get token count, clarity score, and identified issues

  2. 2

    Reviews issues: redundant context, missing output format specification, ambiguous instruction verbs

  3. 3

    Applies patterns: removes redundant lines, adds explicit JSON schema, rewrites vague verbs to concrete actions

  4. 4

    Runs --optimize to generate the revised prompt file

  5. 5

    Compares optimized vs baseline with --compare to show token reduction and clarity improvement

ВЫВОД

Optimized prompt file with a comparison report showing token reduction percentage and resolved issues list

Установка

clawhub install alirezarezvani/senior-prompt-engineer

Выполните команду CLI или вставьте промпт в Claude Code, Cursor, Codex — любой кодинг-агент, которым пользуетесь.

Ссылки