REPOGEO 报告 · LITE
keirp/automatic_prompt_engineer
默认分支 main · commit eac521c7 · 扫描时间 2026/5/11 12:58:19
星标 1,355 · Fork 172
行动计划告诉你下一步要做什么——按影响力排序、可直接复制粘贴的修改。品类可见性是真正的 GEO 测试:当用户向 AI 提一个不带品牌、本应让 keirp/automatic_prompt_engineer 浮出水面的问题时,AI 是真的推荐了你,还是推荐了你的竞品?客观检查验证 AI 引擎最先权衡的那些元数据信号。自指检查判断 AI 是否还认识你的名字。
行动计划 — 可复制粘贴的修复
2 条由 gemini-2.5-flash 生成、按优先级排序的修改。修完后请把对应条目标记为完成。
- highabout#1Add a concise repository description
原因:
复制粘贴的修复Automates the discovery and optimization of prompts for Large Language Models (LLMs) using a meta-LLM approach.
- mediumreadme#2Reposition README opening to highlight the tool's function
原因:
当前# Large Language Models Are Human-Level Prompt Engineers Yongchao Zhou*, Andrei Ioan Muresanu*, Ziwen Han*, Keiran Paster, Silviu Pitis, Harris Chan, Jimmy Ba Project Page | ArXiv | Colab | Demo This repo contains code for "Large Language Models Are Human-Level Prompt Engineers". Please see our paper and project page for more results. # Abstract
复制粘贴的修复# Automatic Prompt Engineer (APE): Automating LLM Prompt Discovery and Optimization This repository provides the code for the Automatic Prompt Engineer (APE) method, which automates the challenging process of discovering and optimizing effective prompts for Large Language Models (LLMs). Inspired by classical program synthesis, APE treats instructions as "programs" optimized by searching over LLM-proposed candidates to maximize a chosen score function. Our method significantly outperforms prior LLM baselines and achieves human-level performance on many NLP tasks. ## Paper & Project Page Yongchao Zhou*, Andrei Ioan Muresanu*, Ziwen Han*, Keiran Paster, Silviu Pitis, Harris Chan, Jimmy Ba Project Page | ArXiv | Colab | Demo This repo contains code for "Large Language Models Are Human-Level Prompt Engineers". Please see our paper and project page for more results. # Abstract
本次扫描解析到的品类 GEO 通道:google/gemini-2.5-flash, deepseek/deepseek-v4-flash
品类可见性 — 真正的 GEO 测试
向 google/gemini-2.5-flash 提出的不带品牌问题。AI 推荐了你,还是推荐了别人?
各模型使用同一组问题 — 切换标签对比回答与排名。
- PromptPerfect · 被推荐 2 次
- langchain-ai/langchain · 被推荐 1 次
- run-llama/llama_index · 被推荐 1 次
- OpenAI API · 被推荐 1 次
- Anthropic · 被推荐 1 次
- 品类问题How to automate the process of prompt engineering for large language models?你:未被推荐AI 推荐顺序:
- LangChain (langchain-ai/langchain)
- LlamaIndex (run-llama/llama_index)
- OpenAI API
- Anthropic
- Jinja2 (pallets/jinja)
- W&B Prompts
- Humanloop (humanloop/humanloop-python)
- Guardrails AI (guardrails-ai/guardrails)
- PromptPerfect
AI 推荐了 10 个替代方案,却始终没点名 keirp/automatic_prompt_engineer。这就是要补上的差距。
查看 AI 完整回答
- 品类问题What methods exist for automatically optimizing prompts to enhance LLM task accuracy?你:未被推荐AI 推荐顺序:
- PromptPerfect
- OpenAI Evals
- RLHF (Reinforcement Learning from Human Feedback)
- Tonic.ai
- Optuna
- DEAP
- Hyperopt
- Scikit-optimize
- AutoPrompt
- LangChain
- LlamaIndex
AI 推荐了 11 个替代方案,却始终没点名 keirp/automatic_prompt_engineer。这就是要补上的差距。
查看 AI 完整回答
客观检查
针对 AI 引擎最看重的元数据信号的规则审计。
- Metadata completenessfail
建议:
- README presencepass
自指检查
当被直接问到你时,AI 是否还知道你的仓库存在?
- Compared to common alternatives in this category, what is the core differentiator of keirp/automatic_prompt_engineer?passAI 未点名 keirp/automatic_prompt_engineer —— 很可能在说另一个项目
AI 的回答可能信誓旦旦却是错的。请按事实核对:技术栈、目标人群、差异化点是不是和你实际的对得上?
- If a team adopts keirp/automatic_prompt_engineer in production, what risks or prerequisites should they evaluate first?passAI 明确点名了 keirp/automatic_prompt_engineer
AI 的回答可能信誓旦旦却是错的。请按事实核对:技术栈、目标人群、差异化点是不是和你实际的对得上?
- In one sentence, what problem does the repo keirp/automatic_prompt_engineer solve, and who is the primary audience?passAI 明确点名了 keirp/automatic_prompt_engineer
AI 的回答可能信誓旦旦却是错的。请按事实核对:技术栈、目标人群、差异化点是不是和你实际的对得上?
嵌入你的 GEO 徽章
把这个徽章贴进 keirp/automatic_prompt_engineer 的 README。每次重新扫描都会自动更新,并跳到最新报告——是「我在乎 AI 可发现性」最简单的公开证明。
[](https://repogeo.com/zh/r/keirp/automatic_prompt_engineer)<a href="https://repogeo.com/zh/r/keirp/automatic_prompt_engineer"><img src="https://repogeo.com/badge/keirp/automatic_prompt_engineer.svg" alt="RepoGEO" /></a>订阅 Pro,解锁深度诊断
keirp/automatic_prompt_engineer — 轻量扫描仍免费;本卡列出 Pro 相对轻量的深度额度。
- 深度报告每月 10 次
- 无品牌品类查询5,轻量 2
- 优先行动项8,轻量 3