应用产品
查找因环境问题(缺少密钥、工具、文件、权限、网络或上下文)导致运行失败的 AI agent,基于 TypeSafe Jev。
Find the AI agent runs that broke because their environment did: missing keys, tools, files, permissions, network, or context. Powered by TypeSafe Jev.
from typesafe_sdk import AsyncTypeSafeClient, TypeSafeError链接指向我们抓取当天的那个 commit,行号是准的。
下面是从这个项目源码里原样取出来的 question 组合。
取自 src/opensmoke/questions.py:55
claims_success是/否The final response presents the task as successfully completed.
import { experimental_evaluate as evaluate } from 'ai';
const { answers } = await evaluate({
model: 'typesafe-ai/jev',
state,
questions: {
claims_success: { type: 'boolean', instructions: 'The final response presents the task as successfully completed.' },
},
});