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:
Ранее Мерц призвал Европу забыть о «наивном пацифизме» в эпоху между войной и миром. Он заявил, что европейские страны, включая ФРГ, должны взять ответственность за оборону собственных границ в свои руки и научиться говорить на языке силы.，详情可参考搜狗输入法下载
В Финляндии предупредили об опасном шаге ЕС против России09:28，推荐阅读heLLoword翻译官方下载获取更多信息
When asked about this, specifically in relation to Styles, Dunstan said, "we are so honoured that out of any venue in the world that Harry and his team have chosen Co-op Live".