Open source file-system and knowledge based agent template. Build AI agents that stay up to date with your knowledge base
import { tool } from 'ai'
import { z } from 'zod'
import { db, schema } from '@nuxthub/db'
import { eq, isNotNull } from 'drizzle-orm'
import { preview, cmd } from './_preview'
export const listUsersTool = tool({
description: `List application users with their usage statistics.
Use this to find active users, check who is using the app, and see per-user token consumption.`,
inputSchema: z.object({
limit: z.number().min(1).max(50).default(20).describe('Maximum number of users to return'),
sortBy: z.enum(['messages', 'tokens', 'recent']).default('messages').describe('Sort users by message count, token usage, or recent activity'),
}),
execute: async function* ({ limit, sortBy }) {
const label = 'List users'
yield { status: 'loading' as const, commands: [cmd(label, '')] }
const start = Date.now()
const users = await db
.select({
id: schema.user.id,
name: schema.user.name,
email: schema.user.email,
image: schema.user.image,
role: schema.user.role,
createdAt: schema.user.createdAt,
})
.from(schema.user)
const userStats = await Promise.all(
users.map(async (user) => {
const userChats = await db
.select({ id: schema.chats.id })
.from(schema.chats)
.where(eq(schema.chats.userId, user.id))
if (userChats.length === 0) {
return { ...user, messageCount: 0, totalInputTokens: 0, totalOutputTokens: 0, chatCount: 0 }
}
const chatIds = userChats.map(c => c.id)
const messages = await db
.select({
chatId: schema.messages.chatId,
inputTokens: schema.messages.inputTokens,
outputTokens: schema.messages.outputTokens,
})
.from(schema.messages)
.where(isNotNull(schema.messages.model))
const userMessages = messages.filter(m => chatIds.includes(m.chatId))
let totalInputTokens = 0
let totalOutputTokens = 0
for (const msg of userMessages) {
totalInputTokens += msg.inputTokens ?? 0
totalOutputTokens += msg.outputTokens ?? 0
}
return {
...user,
messageCount: userMessages.length,
totalInputTokens,
totalOutputTokens,
chatCount: userChats.length,
}
}),
)
if (sortBy === 'tokens') {
userStats.sort((a, b) => (b.totalInputTokens + b.totalOutputTokens) - (a.totalInputTokens + a.totalOutputTokens))
} else if (sortBy === 'recent') {
userStats.sort((a, b) => new Date(b.createdAt).getTime() - new Date(a.createdAt).getTime())
} else {
userStats.sort((a, b) => b.messageCount - a.messageCount)
}
const sliced = userStats.slice(0, limit)
const usersData = sliced.map(u => ({
id: u.id,
name: u.name,
email: u.email,
role: u.role,
createdAt: u.createdAt,
chatCount: u.chatCount,
messageCount: u.messageCount,
totalTokens: u.totalInputTokens + u.totalOutputTokens,
}))
yield {
status: 'done' as const,
commands: [cmd(label, preview(usersData))],
durationMs: Date.now() - start,
users: sliced.map(u => ({
id: u.id,
name: u.name,
email: u.email,
role: u.role,
createdAt: u.createdAt,
chatCount: u.chatCount,
messageCount: u.messageCount,
totalTokens: u.totalInputTokens + u.totalOutputTokens,
})),
}
},
})