free-coding-models 0.3.9 → 0.3.12

package/src/proxy-server.js

@@ -1,1477 +0,0 @@
-/**
- * @file lib/proxy-server.js
- * @description Multi-account rotation proxy server with SSE streaming,
- * token stats tracking, Anthropic/OpenAI translation, and persistent request logging.
- *
- * Design:
- *   - Binds to 127.0.0.1 only (never 0.0.0.0)
- *   - SSE is piped through without buffering (upstreamRes.pipe(clientRes))
- *   - HTTP/HTTPS module is chosen BEFORE the request is created (single code-path)
- *   - x-ratelimit-* headers are stripped from all responses forwarded to clients
- *   - Retry loop: first attempt uses sticky session fingerprint; subsequent
- *     retries use fresh P2C to avoid hitting the same failed account
- *   - Claude-family aliases are resolved inside the proxy so Claude Code can
- *     keep emitting `claude-*` / `sonnet` / `haiku` style model ids safely
- *
- * @exports ProxyServer
- */
-
-import http from 'node:http'
-import https from 'node:https'
-import { AccountManager } from './account-manager.js'
-import { classifyError } from './error-classifier.js'
-import { applyThinkingBudget, compressContext } from './request-transformer.js'
-import { TokenStats } from './token-stats.js'
-import { createHash } from 'node:crypto'
-import {
-  translateAnthropicToOpenAI,
-  translateOpenAIToAnthropic,
-  createAnthropicSSETransformer,
-  estimateAnthropicTokens,
-} from './anthropic-translator.js'
-import {
-  translateResponsesToOpenAI,
-  translateOpenAIToResponses,
-  createResponsesSSETransformer,
-} from './responses-translator.js'
-
-// ─── Helpers ─────────────────────────────────────────────────────────────────
-
-/**
- * Choose the http or https module based on the URL scheme.
- * MUST be called before creating the request (single code-path).
- *
- * @param {string} url
- * @returns {typeof import('http') | typeof import('https')}
- */
-function selectClient(url) {
-  return url.startsWith('https') ? https : http
-}
-
-/**
- * Return a copy of the headers object with all x-ratelimit-* entries removed.
- *
- * @param {Record<string, string | string[]>} headers
- * @returns {Record<string, string | string[]>}
- */
-function stripRateLimitHeaders(headers) {
-  const result = {}
-  for (const [key, value] of Object.entries(headers)) {
-    if (!key.toLowerCase().startsWith('x-ratelimit')) {
-      result[key] = value
-    }
-  }
-  return result
-}
-
-// 📖 Max body size limit to prevent memory exhaustion attacks (10 MB)
-const MAX_BODY_SIZE = 10 * 1024 * 1024
-
-/**
- * Buffer all chunks from an http.IncomingMessage and return the body as a string.
- * Enforces a size limit to prevent memory exhaustion from oversized payloads.
- *
- * @param {http.IncomingMessage} req
- * @returns {Promise<string>}
- * @throws {Error} with statusCode 413 if body exceeds MAX_BODY_SIZE
- */
-function readBody(req) {
-  return new Promise((resolve, reject) => {
-    const chunks = []
-    let totalSize = 0
-    req.on('data', chunk => {
-      totalSize += chunk.length
-      if (totalSize > MAX_BODY_SIZE) {
-        req.destroy()
-        const err = new Error('Request body too large')
-        err.statusCode = 413
-        return reject(err)
-      }
-      chunks.push(chunk)
-    })
-    req.on('end', () => resolve(Buffer.concat(chunks).toString()))
-    req.on('error', reject)
-  })
-}
-
-/**
- * Write a JSON (or pre-serialised) response to the client.
- *
- * @param {http.ServerResponse} res
- * @param {number} statusCode
- * @param {object | string} body
- */
-function sendJson(res, statusCode, body) {
-  if (res.headersSent) return
-  const json = typeof body === 'string' ? body : JSON.stringify(body)
-  res.writeHead(statusCode, { 'content-type': 'application/json' })
-  res.end(json)
-}
-
-/**
- * 📖 Match routes on the URL pathname only so Claude Code's `?beta=true`
- * 📖 Anthropic requests resolve exactly like FastAPI routes do in free-claude-code.
- *
- * @param {http.IncomingMessage} req
- * @returns {string}
- */
-function getRequestPathname(req) {
-  try {
-    return new URL(req.url || '/', 'http://127.0.0.1').pathname || '/'
-  } catch {
-    return req.url || '/'
-  }
-}
-
-function normalizeRequestedModel(modelId) {
-  if (typeof modelId !== 'string') return null
-  const trimmed = modelId.trim()
-  if (!trimmed) return null
-  return trimmed.replace(/^fcm-proxy\//, '')
-}
-
-function normalizeAnthropicRouting(anthropicRouting = null) {
-  return {
-    model: normalizeRequestedModel(anthropicRouting?.model),
-    modelOpus: normalizeRequestedModel(anthropicRouting?.modelOpus),
-    modelSonnet: normalizeRequestedModel(anthropicRouting?.modelSonnet),
-    modelHaiku: normalizeRequestedModel(anthropicRouting?.modelHaiku),
-  }
-}
-
-function classifyClaudeVirtualModel(modelId) {
-  const normalized = normalizeRequestedModel(modelId)
-  if (!normalized) return null
-
-  const lower = normalized.toLowerCase()
-
-  // 📖 Mirror free-claude-code's family routing approach: classify by Claude
-  // 📖 family keywords, not only exact ids. Claude Code regularly emits both
-  // 📖 short aliases (`sonnet`) and full versioned ids (`claude-3-5-sonnet-*`).
-  if (lower === 'default') return 'default'
-  if (/^opus(?:plan)?(?:\[1m\])?$/.test(lower)) return 'opus'
-  if (/^sonnet(?:\[1m\])?$/.test(lower)) return 'sonnet'
-  if (lower === 'haiku') return 'haiku'
-  if (!lower.startsWith('claude-')) return null
-  if (lower.includes('opus')) return 'opus'
-  if (lower.includes('haiku')) return 'haiku'
-  if (lower.includes('sonnet')) return 'sonnet'
-  return null
-}
-
-function resolveAnthropicMappedModel(modelId, anthropicRouting) {
-  const routing = normalizeAnthropicRouting(anthropicRouting)
-  const fallbackModel = routing.model
-  if (!fallbackModel && !routing.modelOpus && !routing.modelSonnet && !routing.modelHaiku) {
-    return null
-  }
-
-  const family = classifyClaudeVirtualModel(modelId)
-  if (family === 'opus') return routing.modelOpus || fallbackModel
-  if (family === 'sonnet') return routing.modelSonnet || fallbackModel
-  if (family === 'haiku') return routing.modelHaiku || fallbackModel
-
-  // 📖 free-claude-code falls back to MODEL for unknown Claude ids too.
-  return fallbackModel
-}
-
-function parseProxyAuthorizationHeader(authorization, expectedToken) {
-  if (!expectedToken) return { authorized: true, modelHint: null }
-  if (typeof authorization !== 'string' || !authorization.startsWith('Bearer ')) {
-    return { authorized: false, modelHint: null }
-  }
-
-  const rawToken = authorization.slice('Bearer '.length).trim()
-  if (rawToken === expectedToken) return { authorized: true, modelHint: null }
-  if (!rawToken.startsWith(`${expectedToken}:`)) return { authorized: false, modelHint: null }
-
-  const modelHint = normalizeRequestedModel(rawToken.slice(expectedToken.length + 1))
-  return modelHint
-    ? { authorized: true, modelHint }
-    : { authorized: false, modelHint: null }
-}
-
-// ─── ProxyServer ─────────────────────────────────────────────────────────────
-
-export class ProxyServer {
-  /**
- * @param {{
- *   port?: number,
- *   accounts?: Array<{ id: string, providerKey: string, apiKey: string, modelId: string, url: string }>,
- *   retries?: number,
- *   proxyApiKey?: string,
- *   anthropicRouting?: { model?: string|null, modelOpus?: string|null, modelSonnet?: string|null, modelHaiku?: string|null },
- *   accountManagerOpts?: object,
- *   tokenStatsOpts?: object,
- *   thinkingConfig?: { mode: string, budget_tokens?: number },
- *   compressionOpts?: { level?: number, toolResultMaxChars?: number, thinkingMaxChars?: number, maxTotalChars?: number },
- *   upstreamTimeoutMs?: number
- * }} opts
- */
-  constructor({
-    port = 0,
-    accounts = [],
-    retries = 3,
-    proxyApiKey = null,
-    anthropicRouting = null,
-    accountManagerOpts = {},
-    tokenStatsOpts = {},
-    thinkingConfig,
-    compressionOpts,
-    upstreamTimeoutMs = 45_000,
-  } = {}) {
-    this._port = port
-    this._retries = retries
-    this._thinkingConfig = thinkingConfig
-    this._compressionOpts = compressionOpts
-    this._proxyApiKey = proxyApiKey
-    this._anthropicRouting = normalizeAnthropicRouting(anthropicRouting)
-    this._accounts = accounts
-    this._upstreamTimeoutMs = upstreamTimeoutMs
-    // 📖 Progressive backoff delays (ms) for retries — first attempt is immediate,
-    // subsequent ones add increasing delay + random jitter (0-100ms) to avoid
-    // re-hitting the same rate-limit window on 429s from providers
-    this._retryDelays = [0, 300, 800]
-    this._accountManager = new AccountManager(accounts, accountManagerOpts)
-    this._tokenStats = new TokenStats(tokenStatsOpts)
-    this._startTime = Date.now()
-    this._running = false
-    this._listeningPort = null
-    this._server = http.createServer((req, res) => this._handleRequest(req, res))
-  }
-
-  /**
-   * Start listening on 127.0.0.1.
-   *
-   * @returns {Promise<{ port: number }>}
-   */
-  start() {
-    return new Promise((resolve, reject) => {
-      this._server.once('error', reject)
-      this._server.listen(this._port, '127.0.0.1', () => {
-        this._server.removeListener('error', reject)
-        this._running = true
-        this._listeningPort = this._server.address().port
-        resolve({ port: this._listeningPort })
-      })
-    })
-  }
-
-  /**
-   * Save stats and close the server.
-   *
-   * @returns {Promise<void>}
-   */
-  stop() {
-    this._tokenStats.save()
-    return new Promise(resolve => {
-      this._server.close(() => {
-        this._running = false
-        this._listeningPort = null
-        resolve()
-      })
-    })
-  }
-
-  getStatus() {
-    return {
-      running: this._running,
-      port: this._listeningPort,
-      accountCount: this._accounts.length,
-      healthByAccount: this._accountManager.getAllHealth(),
-      anthropicRouting: this._anthropicRouting,
-    }
-  }
-
-  _getAuthContext(req) {
-    return parseProxyAuthorizationHeader(req.headers.authorization, this._proxyApiKey)
-  }
-
-  _isAuthorized(req) {
-    return this._getAuthContext(req).authorized
-  }
-
-  _resolveAnthropicRequestedModel(modelId, authModelHint = null) {
-    const requestedModel = normalizeRequestedModel(modelId)
-    if (requestedModel && this._accountManager.hasAccountsForModel(requestedModel)) {
-      return requestedModel
-    }
-
-    const mappedModel = resolveAnthropicMappedModel(requestedModel, this._anthropicRouting)
-    if (mappedModel && this._accountManager.hasAccountsForModel(mappedModel)) {
-      return mappedModel
-    }
-
-    // 📖 Claude Code still emits internal aliases / tier model ids for some
-    // 📖 background and helper paths. Keep the old auth-token hint as a final
-    // 📖 compatibility fallback for already-launched sessions, but the primary
-    // 📖 routing path is now the free-claude-code style proxy-side mapping above.
-    if (authModelHint && this._accountManager.hasAccountsForModel(authModelHint)) {
-      if (!requestedModel || classifyClaudeVirtualModel(requestedModel) || requestedModel.toLowerCase().startsWith('claude-')) {
-        return authModelHint
-      }
-    }
-
-    return requestedModel
-  }
-
-  // ── Request routing ────────────────────────────────────────────────────────
-
-  _handleRequest(req, res) {
-    const pathname = getRequestPathname(req)
-
-    // 📖 Root endpoint is unauthenticated so a browser hit on http://127.0.0.1:{port}/
-    // 📖 gives a useful status payload instead of a misleading Unauthorized error.
-    if (req.method === 'GET' && pathname === '/') {
-      return this._handleRoot(res)
-    }
-
-    // 📖 Health endpoint is unauthenticated so external monitors can probe it
-    if (req.method === 'GET' && pathname === '/v1/health') {
-      return this._handleHealth(res)
-    }
-
-    const authContext = this._getAuthContext(req)
-    if (!authContext.authorized) {
-      return sendJson(res, 401, { error: 'Unauthorized' })
-    }
-
-    if (req.method === 'GET' && pathname === '/v1/models') {
-      this._handleModels(res)
-    } else if (req.method === 'GET' && pathname === '/v1/stats') {
-      this._handleStats(res)
-    } else if (req.method === 'POST' && pathname === '/v1/chat/completions') {
-      this._handleChatCompletions(req, res).catch(err => {
-        console.error('[proxy] Internal error:', err)
-        // 📖 Return 413 for body-too-large, generic 500 for everything else — never leak stack traces
-        const status = err.statusCode === 413 ? 413 : 500
-        const msg = err.statusCode === 413 ? 'Request body too large' : 'Internal server error'
-        sendJson(res, status, { error: msg })
-      })
-    } else if (req.method === 'POST' && pathname === '/v1/messages') {
-      // 📖 Anthropic Messages API translation — enables Claude Code compatibility
-      this._handleAnthropicMessages(req, res, authContext).catch(err => {
-        console.error('[proxy] Internal error:', err)
-        const status = err.statusCode === 413 ? 413 : 500
-        const msg = err.statusCode === 413 ? 'Request body too large' : 'Internal server error'
-        sendJson(res, status, { error: msg })
-      })
-    } else if (req.method === 'POST' && pathname === '/v1/messages/count_tokens') {
-      this._handleAnthropicCountTokens(req, res).catch(err => {
-        console.error('[proxy] Internal error:', err)
-        const status = err.statusCode === 413 ? 413 : 500
-        const msg = err.statusCode === 413 ? 'Request body too large' : 'Internal server error'
-        sendJson(res, status, { error: msg })
-      })
-    } else if (req.method === 'POST' && pathname === '/v1/responses') {
-      this._handleResponses(req, res).catch(err => {
-        console.error('[proxy] Internal error:', err)
-        const status = err.statusCode === 413 ? 413 : 500
-        const msg = err.statusCode === 413 ? 'Request body too large' : 'Internal server error'
-        sendJson(res, status, { error: msg })
-      })
-    } else if (req.method === 'POST' && pathname === '/v1/completions') {
-      // These legacy/alternative OpenAI endpoints are not supported by the proxy.
-      // Return 501 (not 404) so callers get a clear signal instead of silently failing.
-      sendJson(res, 501, {
-        error: 'Not Implemented',
-        message: `${pathname} is not supported by this proxy. Use POST /v1/chat/completions instead.`,
-      })
-    } else {
-      sendJson(res, 404, { error: 'Not found' })
-    }
-  }
-
-  // ── GET /v1/models ─────────────────────────────────────────────────────────
-
-  _handleModels(res) {
-    const seen = new Set()
-    const data = []
-    const models = []
-    for (const acct of this._accounts) {
-      const publicModelId = acct.proxyModelId || acct.modelId
-      if (!seen.has(publicModelId)) {
-        seen.add(publicModelId)
-        const modelEntry = {
-          id: publicModelId,
-          slug: publicModelId,
-          name: publicModelId,
-          object: 'model',
-          created: Math.floor(Date.now() / 1000),
-          owned_by: 'proxy',
-        }
-        data.push(modelEntry)
-        models.push(modelEntry)
-      }
-    }
-    sendJson(res, 200, { object: 'list', data, models })
-  }
-
-  // ── POST /v1/chat/completions ──────────────────────────────────────────────
-
-  async _handleChatCompletions(clientReq, clientRes) {
-    // 1. Read and parse request body
-    const rawBody = await readBody(clientReq)
-    let body
-    try {
-      body = JSON.parse(rawBody)
-    } catch {
-      return sendJson(clientRes, 400, { error: 'Invalid JSON body' })
-    }
-
-    // 2. Optional transformations (both functions return new objects, no mutation)
-    if (this._compressionOpts && Array.isArray(body.messages)) {
-      body = { ...body, messages: compressContext(body.messages, this._compressionOpts) }
-    }
-    if (this._thinkingConfig) {
-      body = applyThinkingBudget(body, this._thinkingConfig)
-    }
-
-    // 3. Session fingerprint for first-attempt sticky routing
-    const fingerprint = createHash('sha256')
-      .update(JSON.stringify(body.messages?.slice(-1) ?? []))
-      .digest('hex')
-      .slice(0, 16)
-
-    const requestedModel = typeof body.model === 'string'
-      ? body.model.replace(/^fcm-proxy\//, '')
-      : undefined
-
-    // 4. Early check: if a specific model is requested but has no registered accounts,
-    // return 404 immediately with a clear message rather than silently failing.
-    if (requestedModel && !this._accountManager.hasAccountsForModel(requestedModel)) {
-      return sendJson(clientRes, 404, {
-        error: 'Model not found',
-        message: `Model '${requestedModel}' is not available through this proxy. Use GET /v1/models to list available models.`,
-      })
-    }
-
-    const formatSwitchReason = (classified) => {
-      switch (classified?.type) {
-        case 'QUOTA_EXHAUSTED':
-          return 'quota'
-        case 'RATE_LIMITED':
-          return '429'
-        case 'MODEL_NOT_FOUND':
-          return '404'
-        case 'MODEL_CAPACITY':
-          return 'capacity'
-        case 'SERVER_ERROR':
-          return '5xx'
-        case 'NETWORK_ERROR':
-          return 'network'
-        default:
-          return 'retry'
-      }
-    }
-
-    // 5. Retry loop with progressive backoff
-    let pendingSwitchReason = null
-    let previousAccount = null
-    for (let attempt = 0; attempt < this._retries; attempt++) {
-      // 📖 Apply backoff delay before retries (first attempt is immediate)
-      const delay = this._retryDelays[Math.min(attempt, this._retryDelays.length - 1)]
-      if (delay > 0) await new Promise(r => setTimeout(r, delay + Math.random() * 100))
-
-      // First attempt: respect sticky session.
-      // Subsequent retries: fresh P2C (don't hammer the same failed account).
-      const selectOpts = attempt === 0
-        ? { sessionFingerprint: fingerprint, requestedModel }
-        : { requestedModel }
-      const account = this._accountManager.selectAccount(selectOpts)
-      if (!account) break // No available accounts → fall through to 503
-
-      const result = await this._forwardRequest(account, body, clientRes, {
-        requestedModel,
-        switched: attempt > 0,
-        switchReason: pendingSwitchReason,
-        switchedFromProviderKey: previousAccount?.providerKey,
-        switchedFromModelId: previousAccount?.modelId,
-      })
-
-      // Response fully sent (success JSON or SSE pipe established)
-      if (result.done) return
-
-      // Error path: classify → record → retry or forward error
-      const { statusCode, responseBody, responseHeaders, networkError } = result
-      const classified = classifyError(
-        networkError ? 0 : statusCode,
-        responseBody || '',
-        responseHeaders || {}
-      )
-
-      this._accountManager.recordFailure(account.id, classified, { providerKey: account.providerKey })
-      if (responseHeaders) {
-        const quotaUpdated = this._accountManager.updateQuota(account.id, responseHeaders)
-        this._persistQuotaSnapshot(account, quotaUpdated)
-      }
-
-      if (!classified.shouldRetry) {
-        // Non-retryable (auth error, unknown) → return upstream response directly
-        return sendJson(
-          clientRes,
-          statusCode || 500,
-          responseBody || JSON.stringify({ error: 'Upstream error' })
-        )
-      }
-      // shouldRetry === true → next attempt
-      pendingSwitchReason = formatSwitchReason(classified)
-      previousAccount = account
-    }
-
-    // All retries consumed, or no accounts available from the start
-    sendJson(clientRes, 503, { error: 'All accounts exhausted or unavailable' })
-  }
-
-  // ── Upstream forwarding ────────────────────────────────────────────────────
-
-  /**
-   * Forward one attempt to the upstream API.
-   *
-   * Resolves with:
-   *   { done: true }
-   *     — The response has been committed to clientRes (success JSON sent, or
-   *       SSE pipe established).  The retry loop must return immediately.
-   *
-   *   { done: false, statusCode, responseBody, responseHeaders, networkError }
-   *     — An error occurred; the retry loop decides whether to retry or give up.
-   *
-   * @param {{ id: string, apiKey: string, modelId: string, url: string }} account
-   * @param {object} body
-   * @param {http.ServerResponse} clientRes
-   * @param {{ requestedModel?: string, switched?: boolean, switchReason?: string|null, switchedFromProviderKey?: string, switchedFromModelId?: string }} [logContext]
-   * @returns {Promise<{ done: boolean }>}
-   */
-  _forwardRequest(account, body, clientRes, logContext = {}) {
-    return new Promise(resolve => {
-      // Replace client-supplied model name with the account's model ID
-      const newBody = { ...body, model: account.modelId }
-      const bodyStr = JSON.stringify(newBody)
-
-      // Build the full upstream URL from the account's base URL
-      const baseUrl = account.url.replace(/\/$/, '')
-      let upstreamUrl
-      try {
-        upstreamUrl = new URL(baseUrl + '/chat/completions')
-      } catch {
-        // 📖 Malformed upstream URL — resolve as network error so retry loop can continue
-        return resolve({ done: false, statusCode: 0, responseBody: 'Invalid upstream URL', networkError: true })
-      }
-
-      // Choose http or https module BEFORE creating the request
-      const client = selectClient(account.url)
-      const startTime = Date.now()
-
-      const requestOptions = {
-        hostname: upstreamUrl.hostname,
-        port: upstreamUrl.port || (upstreamUrl.protocol === 'https:' ? 443 : 80),
-        path: upstreamUrl.pathname + (upstreamUrl.search || ''),
-        method: 'POST',
-        headers: {
-          'authorization': `Bearer ${account.apiKey}`,
-          'content-type': 'application/json',
-          'content-length': Buffer.byteLength(bodyStr),
-        },
-      }
-
-      const upstreamReq = client.request(requestOptions, upstreamRes => {
-        const { statusCode } = upstreamRes
-        const headers = upstreamRes.headers
-        const contentType = headers['content-type'] || ''
-        const isSSE = contentType.includes('text/event-stream')
-
-        if (statusCode >= 200 && statusCode < 300) {
-          if (isSSE) {
-            // ── SSE passthrough: MUST NOT buffer ──────────────────────────
-            const strippedHeaders = stripRateLimitHeaders(headers)
-            clientRes.writeHead(statusCode, {
-              ...strippedHeaders,
-              'content-type': 'text/event-stream',
-              'cache-control': 'no-cache',
-            })
-
-            // Tap the data stream to capture usage from the last data line.
-            // Register BEFORE pipe() so both listeners share the same event queue.
-            // 📖 sseLineBuffer persists between chunks to handle lines split across boundaries
-            let lastChunkData = ''
-            let sseLineBuffer = ''
-            upstreamRes.on('data', chunk => {
-              sseLineBuffer += chunk.toString()
-              const lines = sseLineBuffer.split('\n')
-              // 📖 Last element may be an incomplete line — keep it for next chunk
-              sseLineBuffer = lines.pop() || ''
-              for (const line of lines) {
-                if (line.startsWith('data: ') && !line.includes('[DONE]')) {
-                  lastChunkData = line.slice(6).trim()
-                }
-              }
-            })
-
-            upstreamRes.on('end', () => {
-              let promptTokens = 0
-              let completionTokens = 0
-              try {
-                const parsed = JSON.parse(lastChunkData)
-                if (parsed.usage) {
-                  promptTokens = parsed.usage.prompt_tokens || 0
-                  completionTokens = parsed.usage.completion_tokens || 0
-                }
-              } catch { /* no usage in stream — ignore */ }
-              // Always record every upstream attempt so the log page shows real requests
-              this._tokenStats.record({
-                accountId: account.id,
-                modelId: account.modelId,
-                providerKey: account.providerKey,
-                statusCode,
-                requestType: 'chat.completions',
-                promptTokens,
-                completionTokens,
-                latencyMs: Date.now() - startTime,
-                success: true,
-                requestedModelId: logContext.requestedModel,
-                switched: logContext.switched === true,
-                switchReason: logContext.switchReason,
-                switchedFromProviderKey: logContext.switchedFromProviderKey,
-                switchedFromModelId: logContext.switchedFromModelId,
-              })
-              this._accountManager.recordSuccess(account.id, Date.now() - startTime)
-              const quotaUpdated = this._accountManager.updateQuota(account.id, headers)
-              this._persistQuotaSnapshot(account, quotaUpdated)
-            })
-
-            // 📖 Error handlers on both sides of the pipe to prevent uncaught errors
-            upstreamRes.on('error', err => { if (!clientRes.destroyed) clientRes.destroy(err) })
-            clientRes.on('error', () => { if (!upstreamRes.destroyed) upstreamRes.destroy() })
-
-            // Pipe after listeners are registered; upstream → client, no buffering
-            upstreamRes.pipe(clientRes)
-
-            // ── Downstream disconnect cleanup ─────────────────────────────
-            // If the client closes its connection mid-stream, destroy the
-            // upstream request and response promptly so we don't hold the
-            // upstream connection open indefinitely.
-            clientRes.on('close', () => {
-              if (!upstreamRes.destroyed) upstreamRes.destroy()
-              if (!upstreamReq.destroyed) upstreamReq.destroy()
-            })
-
-            // The pipe handles the rest asynchronously; signal done to retry loop
-            resolve({ done: true })
-          } else {
-            // ── JSON response ─────────────────────────────────────────────
-            const chunks = []
-            upstreamRes.on('data', chunk => chunks.push(chunk))
-            upstreamRes.on('end', () => {
-              const responseBody = Buffer.concat(chunks).toString()
-              const latencyMs = Date.now() - startTime
-
-              const quotaUpdated = this._accountManager.updateQuota(account.id, headers)
-              this._accountManager.recordSuccess(account.id, latencyMs)
-              this._persistQuotaSnapshot(account, quotaUpdated)
-
-              // Always record every upstream attempt so the log page shows real requests.
-              // Extract tokens if upstream provides them; default to 0 when not present.
-              let promptTokens = 0
-              let completionTokens = 0
-              try {
-                const parsed = JSON.parse(responseBody)
-                if (parsed.usage) {
-                  promptTokens = parsed.usage.prompt_tokens || 0
-                  completionTokens = parsed.usage.completion_tokens || 0
-                }
-              } catch { /* non-JSON body — tokens stay 0 */ }
-              this._tokenStats.record({
-                accountId: account.id,
-                modelId: account.modelId,
-                providerKey: account.providerKey,
-                statusCode,
-                requestType: 'chat.completions',
-                promptTokens,
-                completionTokens,
-                latencyMs,
-                success: true,
-                requestedModelId: logContext.requestedModel,
-                switched: logContext.switched === true,
-                switchReason: logContext.switchReason,
-                switchedFromProviderKey: logContext.switchedFromProviderKey,
-                switchedFromModelId: logContext.switchedFromModelId,
-              })
-
-              // Forward stripped response to client
-              const strippedHeaders = stripRateLimitHeaders(headers)
-              clientRes.writeHead(statusCode, {
-                ...strippedHeaders,
-                'content-type': 'application/json',
-              })
-              clientRes.end(responseBody)
-              resolve({ done: true })
-            })
-          }
-        } else {
-          // ── Error response: buffer for classification in retry loop ─────
-          const chunks = []
-          upstreamRes.on('data', chunk => chunks.push(chunk))
-          upstreamRes.on('end', () => {
-            const latencyMs = Date.now() - startTime
-            // Log every failed upstream attempt so the log page shows real requests
-            this._tokenStats.record({
-              accountId: account.id,
-              modelId: account.modelId,
-              providerKey: account.providerKey,
-              statusCode,
-              requestType: 'chat.completions',
-              promptTokens: 0,
-              completionTokens: 0,
-              latencyMs,
-              success: false,
-              requestedModelId: logContext.requestedModel,
-              switched: logContext.switched === true,
-              switchReason: logContext.switchReason,
-              switchedFromProviderKey: logContext.switchedFromProviderKey,
-              switchedFromModelId: logContext.switchedFromModelId,
-            })
-            resolve({
-              done: false,
-              statusCode,
-              responseBody: Buffer.concat(chunks).toString(),
-              responseHeaders: headers,
-              networkError: false,
-            })
-          })
-        }
-      })
-
-      upstreamReq.on('error', err => {
-        // TCP / DNS / timeout errors — log as network failure
-        const latencyMs = Date.now() - startTime
-        this._tokenStats.record({
-          accountId: account.id,
-          modelId: account.modelId,
-          providerKey: account.providerKey,
-          statusCode: 0,
-          requestType: 'chat.completions',
-          promptTokens: 0,
-          completionTokens: 0,
-          latencyMs,
-          success: false,
-          requestedModelId: logContext.requestedModel,
-          switched: logContext.switched === true,
-          switchReason: logContext.switchReason,
-          switchedFromProviderKey: logContext.switchedFromProviderKey,
-          switchedFromModelId: logContext.switchedFromModelId,
-        })
-        // TCP / DNS / timeout errors
-        resolve({
-          done: false,
-          statusCode: 0,
-          responseBody: err.message,
-          responseHeaders: {},
-          networkError: true,
-        })
-      })
-
-      // Abort the upstream request if it exceeds the configured timeout.
-      // This prevents indefinite hangs (e.g. nvidia returning 504 after 302 s).
-      // The 'timeout' event fires but does NOT automatically abort; we must call destroy().
-      upstreamReq.setTimeout(this._upstreamTimeoutMs, () => {
-        upstreamReq.destroy(new Error(`Upstream request timed out after ${this._upstreamTimeoutMs}ms`))
-      })
-
-      upstreamReq.write(bodyStr)
-      upstreamReq.end()
-    })
-  }
-
-  /**
-   * Persist a quota snapshot for the given account into TokenStats.
-   * Called after every `AccountManager.updateQuota()` so TUI can read fresh data.
-   * Never exposes apiKey.
-   *
-   * @param {{ id: string, providerKey?: string, modelId?: string }} account
-   * @param {boolean} quotaUpdated
-   */
-  _persistQuotaSnapshot(account, quotaUpdated = true) {
-    if (!quotaUpdated) return
-    const health = this._accountManager.getHealth(account.id)
-    if (!health) return
-    this._tokenStats.updateQuotaSnapshot(account.id, {
-      quotaPercent: health.quotaPercent,
-      ...(account.providerKey !== undefined && { providerKey: account.providerKey }),
-      ...(account.modelId !== undefined && { modelId: account.modelId }),
-    })
-  }
-
-  // ── GET /v1/health ──────────────────────────────────────────────────────────
-
-  /**
-   * 📖 Friendly unauthenticated landing endpoint for browsers and quick local checks.
-   */
-  _handleRoot(res) {
-    const status = this.getStatus()
-    const uniqueModels = new Set(this._accounts.map(acct => acct.proxyModelId || acct.modelId)).size
-    sendJson(res, 200, {
-      status: 'ok',
-      service: 'fcm-proxy-v2',
-      running: status.running,
-      accountCount: status.accountCount,
-      modelCount: uniqueModels,
-      endpoints: {
-        health: '/v1/health',
-        models: '/v1/models',
-        stats: '/v1/stats',
-      },
-    })
-  }
-
-  /**
-   * 📖 Health endpoint for daemon liveness checks. Unauthenticated so external
-   * monitors (TUI, launchctl, systemd) can probe without needing the token.
-   */
-  _handleHealth(res) {
-    const status = this.getStatus()
-    sendJson(res, 200, {
-      status: 'ok',
-      uptime: process.uptime(),
-      port: status.port,
-      accountCount: status.accountCount,
-      running: status.running,
-    })
-  }
-
-  // ── GET /v1/stats ──────────────────────────────────────────────────────────
-
-  /**
-   * 📖 Authenticated stats endpoint — returns per-account health, token stats summary,
-   * and proxy uptime. Useful for monitoring and debugging.
-   */
-  _handleStats(res) {
-    const healthByAccount = this._accountManager.getAllHealth()
-    const summary = this._tokenStats.getSummary()
-
-    // 📖 Compute totals from the summary data
-    const dailyEntries = Object.values(summary.daily || {})
-    const totalRequests = dailyEntries.reduce((sum, d) => sum + (d.requests || 0), 0)
-    const totalTokens = dailyEntries.reduce((sum, d) => sum + (d.tokens || 0), 0)
-
-    sendJson(res, 200, {
-      accounts: healthByAccount,
-      tokenStats: {
-        byModel: summary.byModel || {},
-        recentRequests: summary.recentRequests || [],
-      },
-      anthropicRouting: this._anthropicRouting,
-      totals: {
-        requests: totalRequests,
-        tokens: totalTokens,
-      },
-      uptime: Math.floor((Date.now() - this._startTime) / 1000),
-    })
-  }
-
-  // ── POST /v1/messages (Anthropic translation) ──────────────────────────────
-
-  /**
-   * 📖 Handle Anthropic Messages API requests by translating to OpenAI format,
-   * forwarding through the existing chat completions handler, then translating
-   * the response back to Anthropic format.
-   *
-   * 📖 This makes Claude Code work natively through the FCM proxy.
-   */
-  async _handleAnthropicMessages(clientReq, clientRes, authContext = { modelHint: null }) {
-    const rawBody = await readBody(clientReq)
-    let anthropicBody
-    try {
-      anthropicBody = JSON.parse(rawBody)
-    } catch {
-      return sendJson(clientRes, 400, { error: { type: 'invalid_request_error', message: 'Invalid JSON body' } })
-    }
-
-    // 📖 Translate Anthropic → OpenAI
-    const openaiBody = translateAnthropicToOpenAI(anthropicBody)
-    const resolvedModel = this._resolveAnthropicRequestedModel(openaiBody.model, authContext.modelHint)
-    if (resolvedModel) openaiBody.model = resolvedModel
-    const isStreaming = openaiBody.stream === true
-
-    if (isStreaming) {
-      // 📖 Streaming mode: pipe through SSE transformer
-      await this._handleAnthropicMessagesStreaming(openaiBody, anthropicBody.model, clientRes)
-    } else {
-      // 📖 JSON mode: forward, translate response, return
-      await this._handleAnthropicMessagesJson(openaiBody, anthropicBody.model, clientRes)
-    }
-  }
-
-  /**
-   * 📖 Count tokens for Anthropic Messages requests without calling upstream.
-   * 📖 Claude Code uses this endpoint for budgeting / UI hints, so a fast local
-   * 📖 estimate is enough to keep the flow working through the proxy.
-   */
-  async _handleAnthropicCountTokens(clientReq, clientRes) {
-    const rawBody = await readBody(clientReq)
-    let anthropicBody
-    try {
-      anthropicBody = JSON.parse(rawBody)
-    } catch {
-      return sendJson(clientRes, 400, { error: { type: 'invalid_request_error', message: 'Invalid JSON body' } })
-    }
-
-    sendJson(clientRes, 200, {
-      input_tokens: estimateAnthropicTokens(anthropicBody),
-    })
-  }
-
-  /**
-   * 📖 Handle OpenAI Responses API requests by translating them to chat
-   * 📖 completions, forwarding through the existing proxy path, then converting
-   * 📖 the result back to the Responses wire format.
-   */
-  async _handleResponses(clientReq, clientRes) {
-    const rawBody = await readBody(clientReq)
-    let responsesBody
-    try {
-      responsesBody = JSON.parse(rawBody)
-    } catch {
-      return sendJson(clientRes, 400, { error: 'Invalid JSON body' })
-    }
-
-    const isStreaming = responsesBody.stream === true || String(clientReq.headers.accept || '').includes('text/event-stream')
-    const openaiBody = translateResponsesToOpenAI({ ...responsesBody, stream: isStreaming })
-
-    if (isStreaming) {
-      await this._handleResponsesStreaming(openaiBody, responsesBody.model, clientRes)
-    } else {
-      await this._handleResponsesJson(openaiBody, responsesBody.model, clientRes)
-    }
-  }
-
-  async _handleResponsesJson(openaiBody, requestModel, clientRes) {
-    const capturedChunks = []
-    let capturedStatusCode = 200
-    let capturedHeaders = {}
-
-    const fakeRes = {
-      headersSent: false,
-      destroyed: false,
-      socket: null,
-      writeHead(statusCode, headers) {
-        capturedStatusCode = statusCode
-        capturedHeaders = headers || {}
-        this.headersSent = true
-      },
-      write(chunk) { capturedChunks.push(chunk) },
-      end(data) {
-        if (data) capturedChunks.push(data)
-      },
-      on() { return this },
-      once() { return this },
-      emit() { return false },
-      destroy() { this.destroyed = true },
-      removeListener() { return this },
-    }
-
-    await this._handleChatCompletionsInternal(openaiBody, fakeRes)
-
-    const responseBody = capturedChunks.join('')
-    if (capturedStatusCode >= 200 && capturedStatusCode < 300) {
-      try {
-        const openaiResponse = JSON.parse(responseBody)
-        const responsesResponse = translateOpenAIToResponses(openaiResponse, requestModel)
-        sendJson(clientRes, 200, responsesResponse)
-      } catch {
-        sendJson(clientRes, capturedStatusCode, responseBody)
-      }
-      return
-    }
-
-    // 📖 Forward upstream-style JSON errors unchanged for OpenAI-compatible clients.
-    sendJson(clientRes, capturedStatusCode, responseBody)
-  }
-
-  async _handleResponsesStreaming(openaiBody, requestModel, clientRes) {
-    const { transform } = createResponsesSSETransformer(requestModel)
-    await this._handleResponsesStreamDirect(openaiBody, clientRes, transform)
-  }
-
-  async _handleResponsesStreamDirect(openaiBody, clientRes, sseTransform) {
-    const fingerprint = createHash('sha256')
-      .update(JSON.stringify(openaiBody.messages?.slice(-1) ?? []))
-      .digest('hex')
-      .slice(0, 16)
-
-    const requestedModel = typeof openaiBody.model === 'string'
-      ? openaiBody.model.replace(/^fcm-proxy\//, '')
-      : undefined
-
-    if (requestedModel && !this._accountManager.hasAccountsForModel(requestedModel)) {
-      return sendJson(clientRes, 404, {
-        error: 'Model not found',
-        message: `Model '${requestedModel}' is not available.`,
-      })
-    }
-
-    sseTransform.pipe(clientRes)
-
-    for (let attempt = 0; attempt < this._retries; attempt++) {
-      const delay = this._retryDelays[Math.min(attempt, this._retryDelays.length - 1)]
-      if (delay > 0) await new Promise(r => setTimeout(r, delay + Math.random() * 100))
-
-      const selectOpts = attempt === 0
-        ? { sessionFingerprint: fingerprint, requestedModel }
-        : { requestedModel }
-      const account = this._accountManager.selectAccount(selectOpts)
-      if (!account) break
-
-      const result = await this._forwardRequestForResponsesStream(account, openaiBody, sseTransform, clientRes)
-      if (result.done) return
-
-      const { statusCode, responseBody, responseHeaders, networkError } = result
-      const classified = classifyError(
-        networkError ? 0 : statusCode,
-        responseBody || '',
-        responseHeaders || {}
-      )
-      this._accountManager.recordFailure(account.id, classified, { providerKey: account.providerKey })
-      if (!classified.shouldRetry) {
-        sseTransform.end()
-        return sendJson(clientRes, statusCode || 500, responseBody || JSON.stringify({ error: 'Upstream error' }))
-      }
-    }
-
-    sseTransform.end()
-    sendJson(clientRes, 503, { error: 'All accounts exhausted or unavailable' })
-  }
-
-  /**
-   * 📖 Handle non-streaming Anthropic Messages by internally dispatching to
-   * chat completions logic and translating the JSON response back.
-   */
-  async _handleAnthropicMessagesJson(openaiBody, requestModel, clientRes) {
-    // 📖 Create a fake request/response pair to capture the OpenAI response
-    const capturedChunks = []
-    let capturedStatusCode = 200
-    let capturedHeaders = {}
-
-    const fakeRes = {
-      headersSent: false,
-      destroyed: false,
-      socket: null,
-      writeHead(statusCode, headers) {
-        capturedStatusCode = statusCode
-        capturedHeaders = headers || {}
-        this.headersSent = true
-      },
-      write(chunk) { capturedChunks.push(chunk) },
-      end(data) {
-        if (data) capturedChunks.push(data)
-      },
-      on() { return this },
-      once() { return this },
-      emit() { return false },
-      destroy() { this.destroyed = true },
-      removeListener() { return this },
-    }
-
-    // 📖 Build a fake IncomingMessage-like with pre-parsed body
-    const fakeReq = {
-      method: 'POST',
-      url: '/v1/chat/completions',
-      headers: { 'content-type': 'application/json' },
-      on(event, cb) {
-        if (event === 'data') cb(Buffer.from(JSON.stringify(openaiBody)))
-        if (event === 'end') cb()
-        return this
-      },
-      removeListener() { return this },
-    }
-
-    // 📖 Use internal handler directly instead of fake request
-    await this._handleChatCompletionsInternal(openaiBody, fakeRes)
-
-    const responseBody = capturedChunks.join('')
-
-    if (capturedStatusCode >= 200 && capturedStatusCode < 300) {
-      try {
-        const openaiResponse = JSON.parse(responseBody)
-        const anthropicResponse = translateOpenAIToAnthropic(openaiResponse, requestModel)
-        sendJson(clientRes, 200, anthropicResponse)
-      } catch {
-        // 📖 Couldn't parse — forward raw
-        sendJson(clientRes, capturedStatusCode, responseBody)
-      }
-    } else {
-      // 📖 Error — wrap in Anthropic error format
-      sendJson(clientRes, capturedStatusCode, {
-        type: 'error',
-        error: { type: 'api_error', message: responseBody },
-      })
-    }
-  }
-
-  /**
-   * 📖 Handle streaming Anthropic Messages by forwarding as streaming OpenAI
-   * chat completions and piping through the SSE translator.
-   */
-  async _handleAnthropicMessagesStreaming(openaiBody, requestModel, clientRes) {
-    // 📖 We need to intercept the SSE response and translate it
-    const { transform, getUsage } = createAnthropicSSETransformer(requestModel)
-
-    let resolveForward
-    const forwardPromise = new Promise(r => { resolveForward = r })
-
-    const fakeRes = {
-      headersSent: false,
-      destroyed: false,
-      socket: null,
-      writeHead(statusCode, headers) {
-        this.headersSent = true
-        if (statusCode >= 200 && statusCode < 300) {
-          // 📖 Write Anthropic SSE headers
-          clientRes.writeHead(200, {
-            'content-type': 'text/event-stream',
-            'cache-control': 'no-cache',
-            'connection': 'keep-alive',
-          })
-        } else {
-          clientRes.writeHead(statusCode, headers)
-        }
-      },
-      write(chunk) { /* SSE data handled via pipe */ },
-      end(data) {
-        if (data && !this.headersSent) {
-          // 📖 Non-streaming error response
-          clientRes.end(data)
-        }
-        resolveForward()
-      },
-      on() { return this },
-      once() { return this },
-      emit() { return false },
-      destroy() { this.destroyed = true },
-      removeListener() { return this },
-    }
-
-    // 📖 Actually we need to pipe the upstream SSE through our transformer.
-    // 📖 The simplest approach: use _handleChatCompletionsInternal with stream=true
-    // 📖 and capture the piped response through our transformer.
-
-    // 📖 For streaming, we go lower level — use the retry loop directly
-    await this._handleAnthropicStreamDirect(openaiBody, requestModel, clientRes, transform)
-  }
-
-  /**
-   * 📖 Direct streaming handler for Anthropic messages.
-   * 📖 Runs the retry loop, pipes upstream SSE through the Anthropic transformer.
-   */
-  async _handleAnthropicStreamDirect(openaiBody, requestModel, clientRes, sseTransform) {
-    const { createHash: _createHash } = await import('node:crypto')
-    const fingerprint = _createHash('sha256')
-      .update(JSON.stringify(openaiBody.messages?.slice(-1) ?? []))
-      .digest('hex')
-      .slice(0, 16)
-
-    const requestedModel = typeof openaiBody.model === 'string'
-      ? openaiBody.model.replace(/^fcm-proxy\//, '')
-      : undefined
-
-    if (requestedModel && !this._accountManager.hasAccountsForModel(requestedModel)) {
-      return sendJson(clientRes, 404, {
-        type: 'error',
-        error: { type: 'not_found_error', message: `Model '${requestedModel}' is not available.` },
-      })
-    }
-
-    // 📖 Pipe the transform to client
-    sseTransform.pipe(clientRes)
-
-    for (let attempt = 0; attempt < this._retries; attempt++) {
-      // 📖 Progressive backoff for retries (same as chat completions)
-      const delay = this._retryDelays[Math.min(attempt, this._retryDelays.length - 1)]
-      if (delay > 0) await new Promise(r => setTimeout(r, delay + Math.random() * 100))
-
-      const selectOpts = attempt === 0
-        ? { sessionFingerprint: fingerprint, requestedModel }
-        : { requestedModel }
-      const account = this._accountManager.selectAccount(selectOpts)
-      if (!account) break
-
-      const result = await this._forwardRequestForAnthropicStream(account, openaiBody, sseTransform, clientRes)
-
-      if (result.done) return
-
-      const { statusCode, responseBody, responseHeaders, networkError } = result
-      const classified = classifyError(
-        networkError ? 0 : statusCode,
-        responseBody || '',
-        responseHeaders || {}
-      )
-      this._accountManager.recordFailure(account.id, classified, { providerKey: account.providerKey })
-      if (!classified.shouldRetry) {
-        sseTransform.end()
-        return sendJson(clientRes, statusCode || 500, {
-          type: 'error',
-          error: { type: 'api_error', message: responseBody || 'Upstream error' },
-        })
-      }
-    }
-
-    sseTransform.end()
-    sendJson(clientRes, 503, {
-      type: 'error',
-      error: { type: 'overloaded_error', message: 'All accounts exhausted or unavailable' },
-    })
-  }
-
-  /**
-   * 📖 Forward a streaming request to upstream and pipe SSE through transform.
-   */
-  _forwardRequestForAnthropicStream(account, body, sseTransform, clientRes) {
-    return new Promise(resolve => {
-      const newBody = { ...body, model: account.modelId, stream: true }
-      const bodyStr = JSON.stringify(newBody)
-      const baseUrl = account.url.replace(/\/$/, '')
-      let upstreamUrl
-      try {
-        upstreamUrl = new URL(baseUrl + '/chat/completions')
-      } catch {
-        return resolve({ done: false, statusCode: 0, responseBody: 'Invalid upstream URL', networkError: true })
-      }
-      const client = selectClient(account.url)
-      const startTime = Date.now()
-
-      const requestOptions = {
-        hostname: upstreamUrl.hostname,
-        port: upstreamUrl.port || (upstreamUrl.protocol === 'https:' ? 443 : 80),
-        path: upstreamUrl.pathname + (upstreamUrl.search || ''),
-        method: 'POST',
-        headers: {
-          'authorization': `Bearer ${account.apiKey}`,
-          'content-type': 'application/json',
-          'content-length': Buffer.byteLength(bodyStr),
-        },
-      }
-
-      const upstreamReq = client.request(requestOptions, upstreamRes => {
-        const { statusCode } = upstreamRes
-
-        if (statusCode >= 200 && statusCode < 300) {
-          // 📖 Write Anthropic SSE headers if not already sent
-          if (!clientRes.headersSent) {
-            clientRes.writeHead(200, {
-              'content-type': 'text/event-stream',
-              'cache-control': 'no-cache',
-            })
-          }
-
-          // 📖 Error handlers on both sides of the pipe to prevent uncaught errors
-          upstreamRes.on('error', err => { if (!clientRes.destroyed) clientRes.destroy(err) })
-          clientRes.on('error', () => { if (!upstreamRes.destroyed) upstreamRes.destroy() })
-
-          // 📖 Pipe upstream SSE through Anthropic translator
-          upstreamRes.pipe(sseTransform, { end: true })
-
-          upstreamRes.on('end', () => {
-            this._accountManager.recordSuccess(account.id, Date.now() - startTime)
-          })
-
-          clientRes.on('close', () => {
-            if (!upstreamRes.destroyed) upstreamRes.destroy()
-            if (!upstreamReq.destroyed) upstreamReq.destroy()
-          })
-
-          resolve({ done: true })
-        } else {
-          const chunks = []
-          upstreamRes.on('data', chunk => chunks.push(chunk))
-          upstreamRes.on('end', () => {
-            resolve({
-              done: false,
-              statusCode,
-              responseBody: Buffer.concat(chunks).toString(),
-              responseHeaders: upstreamRes.headers,
-              networkError: false,
-            })
-          })
-        }
-      })
-
-      upstreamReq.on('error', err => {
-        resolve({
-          done: false,
-          statusCode: 0,
-          responseBody: err.message,
-          responseHeaders: {},
-          networkError: true,
-        })
-      })
-
-      upstreamReq.setTimeout(this._upstreamTimeoutMs, () => {
-        upstreamReq.destroy(new Error(`Upstream request timed out after ${this._upstreamTimeoutMs}ms`))
-      })
-
-      upstreamReq.write(bodyStr)
-      upstreamReq.end()
-    })
-  }
-
-  /**
-   * 📖 Forward a streaming chat-completions request and translate the upstream
-   * 📖 SSE stream into Responses API events on the fly.
-   */
-  _forwardRequestForResponsesStream(account, body, sseTransform, clientRes) {
-    return new Promise(resolve => {
-      const newBody = { ...body, model: account.modelId, stream: true }
-      const bodyStr = JSON.stringify(newBody)
-      const baseUrl = account.url.replace(/\/$/, '')
-      let upstreamUrl
-      try {
-        upstreamUrl = new URL(baseUrl + '/chat/completions')
-      } catch {
-        return resolve({ done: false, statusCode: 0, responseBody: 'Invalid upstream URL', networkError: true })
-      }
-
-      const client = selectClient(account.url)
-      const startTime = Date.now()
-      const requestOptions = {
-        hostname: upstreamUrl.hostname,
-        port: upstreamUrl.port || (upstreamUrl.protocol === 'https:' ? 443 : 80),
-        path: upstreamUrl.pathname + (upstreamUrl.search || ''),
-        method: 'POST',
-        headers: {
-          'authorization': `Bearer ${account.apiKey}`,
-          'content-type': 'application/json',
-          'content-length': Buffer.byteLength(bodyStr),
-        },
-      }
-
-      const upstreamReq = client.request(requestOptions, upstreamRes => {
-        const { statusCode } = upstreamRes
-
-        if (statusCode >= 200 && statusCode < 300) {
-          if (!clientRes.headersSent) {
-            clientRes.writeHead(200, {
-              'content-type': 'text/event-stream',
-              'cache-control': 'no-cache',
-            })
-          }
-
-          upstreamRes.on('error', err => { if (!clientRes.destroyed) clientRes.destroy(err) })
-          clientRes.on('error', () => { if (!upstreamRes.destroyed) upstreamRes.destroy() })
-
-          upstreamRes.pipe(sseTransform, { end: true })
-          upstreamRes.on('end', () => {
-            this._accountManager.recordSuccess(account.id, Date.now() - startTime)
-          })
-
-          clientRes.on('close', () => {
-            if (!upstreamRes.destroyed) upstreamRes.destroy()
-            if (!upstreamReq.destroyed) upstreamReq.destroy()
-          })
-
-          resolve({ done: true })
-        } else {
-          const chunks = []
-          upstreamRes.on('data', chunk => chunks.push(chunk))
-          upstreamRes.on('end', () => {
-            resolve({
-              done: false,
-              statusCode,
-              responseBody: Buffer.concat(chunks).toString(),
-              responseHeaders: upstreamRes.headers,
-              networkError: false,
-            })
-          })
-        }
-      })
-
-      upstreamReq.on('error', err => {
-        resolve({
-          done: false,
-          statusCode: 0,
-          responseBody: err.message,
-          responseHeaders: {},
-          networkError: true,
-        })
-      })
-
-      upstreamReq.setTimeout(this._upstreamTimeoutMs, () => {
-        upstreamReq.destroy(new Error(`Upstream request timed out after ${this._upstreamTimeoutMs}ms`))
-      })
-
-      upstreamReq.write(bodyStr)
-      upstreamReq.end()
-    })
-  }
-
-  /**
-   * 📖 Internal version of chat completions handler that takes a pre-parsed body.
-   * 📖 Used by the Anthropic JSON translation path to avoid re-parsing.
-   */
-  async _handleChatCompletionsInternal(body, clientRes) {
-    // 📖 Reuse the exact same logic as _handleChatCompletions but with pre-parsed body
-    if (this._compressionOpts && Array.isArray(body.messages)) {
-      body = { ...body, messages: compressContext(body.messages, this._compressionOpts) }
-    }
-    if (this._thinkingConfig) {
-      body = applyThinkingBudget(body, this._thinkingConfig)
-    }
-
-    const fingerprint = createHash('sha256')
-      .update(JSON.stringify(body.messages?.slice(-1) ?? []))
-      .digest('hex')
-      .slice(0, 16)
-
-    const requestedModel = typeof body.model === 'string'
-      ? body.model.replace(/^fcm-proxy\//, '')
-      : undefined
-
-    if (requestedModel && !this._accountManager.hasAccountsForModel(requestedModel)) {
-      return sendJson(clientRes, 404, {
-        error: 'Model not found',
-        message: `Model '${requestedModel}' is not available.`,
-      })
-    }
-
-    for (let attempt = 0; attempt < this._retries; attempt++) {
-      const delay = this._retryDelays[Math.min(attempt, this._retryDelays.length - 1)]
-      if (delay > 0) await new Promise(r => setTimeout(r, delay + Math.random() * 100))
-
-      const selectOpts = attempt === 0
-        ? { sessionFingerprint: fingerprint, requestedModel }
-        : { requestedModel }
-      const account = this._accountManager.selectAccount(selectOpts)
-      if (!account) break
-
-      const result = await this._forwardRequest(account, body, clientRes, { requestedModel })
-      if (result.done) return
-
-      const { statusCode, responseBody, responseHeaders, networkError } = result
-      const classified = classifyError(
-        networkError ? 0 : statusCode,
-        responseBody || '',
-        responseHeaders || {}
-      )
-      this._accountManager.recordFailure(account.id, classified, { providerKey: account.providerKey })
-      if (!classified.shouldRetry) {
-        return sendJson(clientRes, statusCode || 500, responseBody || JSON.stringify({ error: 'Upstream error' }))
-      }
-    }
-
-    sendJson(clientRes, 503, { error: 'All accounts exhausted or unavailable' })
-  }
-
-  // ── Hot-reload accounts ─────────────────────────────────────────────────────
-
-  /**
-   * 📖 Atomically swap the account list and rebuild the AccountManager.
-   * 📖 Used by the daemon when config changes (new API keys, providers toggled).
-   * 📖 In-flight requests on old accounts will finish naturally.
-   *
-   * @param {Array} accounts — new account list
-   * @param {{ model?: string|null, modelOpus?: string|null, modelSonnet?: string|null, modelHaiku?: string|null }} anthropicRouting
-   */
-  updateAccounts(accounts, anthropicRouting = this._anthropicRouting) {
-    this._accounts = accounts
-    this._anthropicRouting = normalizeAnthropicRouting(anthropicRouting)
-    this._accountManager = new AccountManager(accounts, {})
-  }
-}