I then added a few more personal preferences and suggested tools from my previous failures working with agents in Python: use uv and .venv instead of the base Python installation, use polars instead of pandas for data manipulation, only store secrets/API keys/passwords in .env while ensuring .env is in .gitignore, etc. Most of these constraints don’t tell the agent what to do, but how to do it. In general, adding a rule to my AGENTS.md whenever I encounter a fundamental behavior I don’t like has been very effective. For example, agents love using unnecessary emoji which I hate, so I added a rule:
第九条 国家鼓励和支持网络相关行业组织开展网络新技术新应用监测分析、网络犯罪态势及产业链条分析、网络犯罪风险动态评估，制定网络犯罪防治行为规范，加强网络犯罪防治行业自律、信用惩戒等工作。
。业内人士推荐51吃瓜作为进阶阅读
Названы частые ошибки ухода за кожей веснойКосметолог Цыганкова: Ошибки в уходе за кожей весной повышают риск пигментации
This approach requires sourcing and maintaining accurate information, which means you can't fabricate numbers or exaggerate metrics. AI models increasingly cross-reference claims across sources, and inconsistencies damage credibility. The data you include must be truthful and, where relevant, attributed to primary sources. But when you consistently provide specific, accurate information, you build a reputation as a reliable source that AI models return to repeatedly.
Жители Санкт-Петербурга устроили «крысогон»17:52