feat: bump supercode-cli version to 0.1.87 and enhance token budget management:

apps/supercode-cli/server/package.json

@@ -1,6 +1,6 @@
 {
   "name": "supercode-cli",
-  "version": "0.1.86",
+  "version": "0.1.87",
   "description": "AI-powered coding agent CLI",
   "main": "dist/main.js",
   "bin": {

apps/supercode-cli/server/src/cli/ai/concentrate-service.ts

@@ -4,7 +4,7 @@ import chalk from "chalk"
 import { recordUsage } from "../../lib/track-usage"
 import { computeCost } from "../../lib/pricing"
 import { isEmptyToolResult, isDeniedToolResult, summarizeToolResult, tcName } from "./tool-result"
-import { checkDailyOpusLimit, incrementDailyOpusCount, checkDailyTokenBudget } from "../../lib/token-budget"
+import { checkDailyOpusLimit, incrementDailyOpusCount } from "../../lib/token-budget"
 
 const HIGH_VALUE_MODELS = ["anthropic/claude-fable-5", "anthropic/claude-opus-4-8", "anthropic/claude-opus-4-7", "openai/gpt-5.5"]
 const OPUS_MODELS = ["anthropic/claude-opus-4-8", "anthropic/claude-opus-4-7"]
@@ -118,7 +118,6 @@ export class ConcentrateService {
       const systemMessages = messages.filter(m => m.role === "system")
       const nonSystemMessages = messages.filter(m => m.role !== "system")
       const system = systemMessages.map(m => m.content).join("\n")
-      await checkDailyTokenBudget()
       if (OPUS_MODELS.includes(this.modelName)) {
         await checkDailyOpusLimit()
         await incrementDailyOpusCount()

apps/supercode-cli/server/src/cli/ai/server-proxy-service.ts

@@ -2,11 +2,27 @@ import { getStoredToken } from "src/lib/token"
 import { zodToJsonSchema } from "zod-to-json-schema"
 import type { ModelMessage, FinishReason, LanguageModelUsage } from "ai"
 import { isEmptyToolResult, isDeniedToolResult, summarizeToolResult } from "./tool-result"
+import { appendProxyUsage } from "src/lib/token-budget"
+import prisma from "src/lib/prisma"
 
 const BASE_URL = process.env.SUPERCODE_SERVER_URL || "https://supercode-8w7e.onrender.com"
 
 const MAX_STEPS = 8
 
+async function getUserIdFromToken(): Promise<string | null> {
+  const token = await getStoredToken()
+  if (!token?.access_token) return null
+  try {
+    const session = await prisma.session.findUnique({
+      where: { token: token.access_token as string },
+      select: { userId: true },
+    })
+    return session?.userId ?? null
+  } catch {
+    return null
+  }
+}
+
 export class ServerProxyService {
   readonly modelName: string
   readonly providerName: string
@@ -96,6 +112,9 @@ export class ServerProxyService {
 
     if (!res.ok) {
       const text = await res.text()
+      if (text.includes("Insufficient Funds") || text.includes("Credit usage at configured limit")) {
+        throw new Error("Cloud AI is warming up. Please try again in a moment.")
+      }
       throw new Error(text || "AI proxy request failed")
     }
 
@@ -181,6 +200,7 @@ export class ServerProxyService {
     const seenStepResults: Array<{ toolName: string; result: string }> = []
     const deniedCounts = new Map<string, number>()
     const toolCallHistory: Array<{ toolName: string; argsKey: string }> = []
+    const userId = await getUserIdFromToken()
 
     // Convert tools to JSON-schema definitions ONCE (while the Zod schemas
     // are still intact) so the server can hand valid parameter schemas to
@@ -436,6 +456,19 @@ export class ServerProxyService {
       }
     }
 
+    if (userId && (usage.totalTokens ?? 0) > 0) {
+      appendProxyUsage({
+        provider: this.providerName,
+        model: this.modelName,
+        inputTokens: usage.inputTokens ?? 0,
+        outputTokens: usage.outputTokens ?? 0,
+        cachedInputTokens: usage.inputTokenDetails?.cacheReadTokens ?? 0,
+        totalTokens: usage.totalTokens ?? 0,
+        userId,
+        timestamp: Date.now(),
+      }).catch((e) => console.error("[proxy-usage] Failed to record:", e?.message))
+    }
+
     return {
       content: accumulatedContent,
       finishReason,
@@ -474,6 +507,9 @@ export class ServerProxyService {
 
     if (!res.ok) {
       const text = await res.text()
+      if (text.includes("Insufficient Funds") || text.includes("Credit usage at configured limit")) {
+        throw new Error("Cloud AI is warming up. Please try again in a moment.")
+      }
       throw new Error(text || "AI proxy generate-object request failed")
     }
 

apps/supercode-cli/server/src/cli/commands/slashCommands/index.ts

@@ -35,7 +35,7 @@ export const COMMANDS = [
   { cmd: "/interact", desc: "Browser interaction via Firecrawl" },
   { cmd: "/crawl", desc: "Crawl a website via Firecrawl" },
   { cmd: "/parse", desc: "Parse a file (PDF, DOC, etc.) via Firecrawl" },
-  { cmd: "/usage", desc: "Show daily token usage and budget for Opus 4.8" },
+  { cmd: "/usage", desc: "Show your daily token usage across all models" },
   { cmd: "/token-limit", desc: "Show token limits and usage per model per day" },
   { cmd: "/clear", desc: "Clear current session messages" },
   { cmd: "/new", desc: "Start a new conversation" },

apps/supercode-cli/server/src/cli/commands/slashCommands/usage.ts

@@ -1,45 +1,72 @@
 import chalk from "chalk"
 import prisma from "src/lib/prisma"
-import { theme, sectionHeader, heavyDivider, formatTokenCount } from "src/cli/utils/tui.ts"
-import { getOrCreateDeviceId } from "src/lib/token-budget"
-
-function todayStart(): Date {
-  const now = new Date()
-  return new Date(Date.UTC(now.getUTCFullYear(), now.getUTCMonth(), now.getUTCDate()))
-}
+import {
+  theme,
+  sectionHeader,
+  heavyDivider,
+  formatTokenCount,
+  progressBar,
+} from "src/cli/utils/tui.ts"
+import { getStoredToken } from "src/lib/token"
+import { getUserDailyUsage } from "src/lib/token-budget"
 
 export async function usageCommand(): Promise<void> {
-  const today = todayStart()
-  const tomorrow = new Date(today.getTime() + 86_400_000)
+  const token = await getStoredToken()
+  if (!token?.access_token) {
+    console.log()
+    console.log(heavyDivider())
+    console.log()
+    console.log(`  ${chalk.hex(theme.muted)("Not authenticated. Run /login first.")}`)
+    console.log()
+    console.log(heavyDivider())
+    console.log()
+    return
+  }
+
+  const session = await prisma.session.findUnique({
+    where: { token: token.access_token as string },
+    include: { user: true },
+  })
+  if (!session || session.expiresAt < new Date()) {
+    console.log()
+    console.log(heavyDivider())
+    console.log()
+    console.log(`  ${chalk.hex(theme.muted)("Session expired. Please re-authenticate.")}`)
+    console.log()
+    console.log(heavyDivider())
+    console.log()
+    return
+  }
+
+  const user = session.user
+  const usage = await getUserDailyUsage(user.id)
 
   const w = Math.min(process.stdout.columns ?? 80, 72)
 
   console.log()
   console.log(heavyDivider())
   console.log()
-  console.log(sectionHeader("Usage History", { accent: "green" }))
-
-  const allUsage = await prisma.usageEvent.groupBy({
-    by: ["model"],
-    where: {
-      createdAt: { gte: new Date(Date.now() - 30 * 86_400_000) },
-    },
-    _sum: { totalTokens: true },
-    _count: { id: true },
-  })
+  console.log(sectionHeader("Daily Usage", { accent: "green" }))
+  console.log(`  ${chalk.hex(theme.muted)(`${user.email || user.name} · resets at midnight UTC`)}`)
+  console.log()
 
-  if (allUsage.length > 0) {
-    for (const row of allUsage) {
-      const tokens = row._sum.totalTokens ?? 0
-      const calls = row._count.id
-      const modelLabel = row.model.length > 30 ? row.model.slice(0, 27) + "..." : row.model
-      console.log(`  ${chalk.hex(theme.greenGlow)(modelLabel.padEnd(32))} ${formatTokenCount(tokens).padStart(8)}  ${chalk.hex(theme.muted)(`${calls} calls`)}`)
-    }
-  } else {
-    console.log(`  ${chalk.hex(theme.muted)("No usage recorded in the last 30 days.")}`)
-  }
+  const pct = Math.min(100, Math.round((usage.used / usage.limit) * 100))
+  const color =
+    usage.remaining <= 10_000 ? theme.red
+    : usage.remaining <= 50_000 ? theme.amber
+    : theme.green
 
+  console.log(
+    `  ${chalk.hex(theme.greenGlow)("Used")}      ${formatTokenCount(usage.used).padStart(8)}  ${progressBar(usage.used, usage.limit, 16)}`
+  )
+  console.log(
+    `  ${chalk.hex(theme.greenGlow)("Limit")}    ${formatTokenCount(usage.limit).padStart(8)}`
+  )
+  console.log(
+    `  ${chalk.hex(color)(chalk.hex(theme.greenGlow)("Remaining"))} ${formatTokenCount(usage.remaining).padStart(8)}`
+  )
   console.log()
+
   console.log(heavyDivider())
   console.log()
 }

apps/supercode-cli/server/src/index.ts

@@ -6,6 +6,7 @@ import prisma from "./lib/prisma"
 import { loadEnvOnce } from "./lib/load-env"
 import { recordUsage } from "./lib/track-usage"
 import { computeCost } from "./lib/pricing"
+import { checkDailyTokenBudget } from "./lib/token-budget"
 import { registerAnalyticsRoutes } from "./routes/analytics"
 import { transcribeAudio } from "./voice/speech"
 import { tmpdir } from "os"
@@ -301,6 +302,11 @@ app.post("/api/ai/chat", async (req, res) => {
       return
     }
 
+    const isByok = provider === "concentrateai" && !!req.body.concentrateAiKey
+    if (!isByok) {
+      await checkDailyTokenBudget(user.id)
+    }
+
     const systemMessages = messages.filter((m: any) => m.role === "system")
     const nonSystemMessages = messages.filter((m: any) => m.role !== "system")
     const system = systemMessages.map((m: any) => m.content).join("\n")
@@ -333,6 +339,7 @@ app.post("/api/ai/chat", async (req, res) => {
           inputTokens, outputTokens, cachedInputTokens, totalTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, cachedInputTokens),
           durationMs: Date.now() - googleStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({ type: "finish", reason: await result.finishReason, usage }) + "\n")
         res.end()
@@ -441,6 +448,7 @@ app.post("/api/ai/chat", async (req, res) => {
           totalTokens: inputTokens + outputTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, 0),
           durationMs: Date.now() - orStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({
           type: "finish", reason: "stop",
@@ -478,6 +486,7 @@ app.post("/api/ai/chat", async (req, res) => {
           inputTokens, outputTokens, cachedInputTokens, totalTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, cachedInputTokens),
           durationMs: Date.now() - mmStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({ type: "finish", reason: await result.finishReason, usage }) + "\n")
         res.end()
@@ -588,6 +597,7 @@ app.post("/api/ai/chat", async (req, res) => {
           totalTokens: inputTokens + outputTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, 0),
           durationMs: Date.now() - nvidiaStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({
           type: "finish", reason: "stop",
@@ -698,6 +708,7 @@ app.post("/api/ai/chat", async (req, res) => {
           totalTokens: inputTokens + outputTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, 0),
           durationMs: Date.now() - mdStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({
           type: "finish", reason: "stop",
@@ -809,6 +820,7 @@ app.post("/api/ai/chat", async (req, res) => {
           totalTokens: inputTokens + outputTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, 0),
           durationMs: Date.now() - orStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({
           type: "finish", reason: "stop",
@@ -974,6 +986,7 @@ app.post("/api/ai/chat", async (req, res) => {
           totalTokens: inputTokens + outputTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, 0),
           durationMs: Date.now() - caStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({
           type: "finish", reason: "stop",
@@ -1133,6 +1146,7 @@ app.post("/api/ai/chat", async (req, res) => {
           totalTokens: inputTokens + outputTokens,
           costUsd: computeCost(modelName, inputTokens, outputTokens, 0),
           durationMs: Date.now() - scStart,
+          userId: user.id,
         })
         res.write(JSON.stringify({
           type: "finish", reason: "stop",
@@ -1149,6 +1163,8 @@ app.post("/api/ai/chat", async (req, res) => {
     const msg = String(error)
     if (msg.includes("insufficient balance") || msg.includes("402")) {
       res.status(402).json({ error: "MiniMax API: insufficient balance. Top up at https://platform.minimax.ai" })
+    } else if (msg.includes("daily limit of 128K tokens")) {
+      res.status(429).json({ error: msg })
     } else {
       res.status(500).json({ error: msg })
     }
@@ -1169,6 +1185,11 @@ app.post("/api/ai/generate-object", async (req, res) => {
       return
     }
 
+    const isByok = provider === "concentrateai" && !!req.body.concentrateAiKey
+    if (!isByok) {
+      await checkDailyTokenBudget(user.id)
+    }
+
     const { generateObject } = await import("ai")
 
     switch (provider) {
@@ -1327,6 +1348,8 @@ app.post("/api/ai/generate-object", async (req, res) => {
     const msg = String(error)
     if (msg.includes("insufficient balance") || msg.includes("402")) {
       res.status(402).json({ error: "MiniMax API: insufficient balance. Top up at https://platform.minimax.ai" })
+    } else if (msg.includes("daily limit of 128K tokens")) {
+      res.status(429).json({ error: msg })
     } else {
       res.status(500).json({ error: msg })
     }