feat: align with CommandCode CLI 1.28.1 and instrument stalled turns

Add Qwen 3.8 27B to the static catalog (56 models), bump the CLI
version header, and log warn-level diagnostics when a streaming turn
dies mid-way or a non-streaming turn ends in an upstream 502, so
intermittent disconnects can be attributed from logs alone.

Author: yelixir-dev

.env.example

@@ -23,7 +23,7 @@ COMMANDCODE_API_BASE=https://api.commandcode.ai
 COMMANDCODE_DEFAULT_MODEL=deepseek/deepseek-v4-pro
 COMMANDCODE_ALLOWED_MODELS=deepseek/deepseek-v4-pro,deepseek/deepseek-v4-flash
 COMMANDCODE_ALLOW_UNKNOWN_MODELS=false
-COMMANDCODE_CLI_VERSION=1.25.0
+COMMANDCODE_CLI_VERSION=1.28.1
 COMMANDCODE_TIMEOUT_MS=600000
 COMMANDCODE_RETRY_MAX_ATTEMPTS=5
 COMMANDCODE_RETRY_BACKOFF_MS=250

README.ko.md

@@ -8,7 +8,7 @@
 </p>
 
 <p align="center">
-  <a href="package.json"><img src="https://img.shields.io/badge/version-1.25.0.b-b57920?style=flat-square" alt="Version 1.25.0.b"></a>
+  <a href="package.json"><img src="https://img.shields.io/badge/version-1.28.1.a-b57920?style=flat-square" alt="Version 1.28.1.a"></a>
   <a href="src/model-catalog.ts"><img src="https://img.shields.io/badge/models-52-1f6f78?style=flat-square" alt="52 models"></a>
   <a href="package.json"><img src="https://img.shields.io/badge/Node.js-20%2B-9f4d2e?style=flat-square" alt="Node.js 20+"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-28231f?style=flat-square" alt="MIT License"></a>
@@ -20,7 +20,7 @@
 
 <!-- README-I18N:END -->
 
-CommandCode Bridge는 CommandCode 계정을 위한 신뢰 환경용 HTTP 게이트웨이입니다. 표준 OpenAI-compatible 모델·채팅 endpoint를 제공하고 eligible upstream credential 사이에서 요청을 라우팅하며, CommandCode **1.25.0**에 맞춘 정확한 **55-model** catalog를 게시합니다. Bridge 버전은 항상 현재 CommandCode CLI 버전을 따라가며 그 뒤에 문자 접미사를 붙입니다(예: **1.25.0.b**). 접미사는 bridge 전용 release를 뜻합니다.
+CommandCode Bridge는 CommandCode 계정을 위한 신뢰 환경용 HTTP 게이트웨이입니다. 표준 OpenAI-compatible 모델·채팅 endpoint를 제공하고 eligible upstream credential 사이에서 요청을 라우팅하며, CommandCode **1.28.1**에 맞춘 정확한 **56-model** catalog를 게시합니다. Bridge 버전은 항상 현재 CommandCode CLI 버전을 따라가며 그 뒤에 문자 접미사를 붙입니다(예: **1.28.1.a**). 접미사는 bridge 전용 release를 뜻합니다.
 
 [기능](#기능) · [설치](#설치) · [사용법](#사용법) · [동작 방식](#동작-방식) · [저장소 구성](#저장소-구성) · [현재 제한](#현재-제한) · [라이선스](#라이선스)
 
@@ -48,7 +48,7 @@ CommandCode Bridge는 CommandCode 계정을 위한 신뢰 환경용 HTTP 게이
 
 ### Linux rootless installer
 
-installer는 Linux user systemd용이며 CommandCode CLI 1.25.0 때문에 Node.js 22+를 요구합니다. 가능한 경우 CLI auth를 가져오고 private state는 `~/.config/commandcode-bridge`, 설치본은 `~/.local/share/commandcode-bridge`에 두며 안전한 기본값은 `127.0.0.1:9992`입니다. `0.0.0.0`은 `BRIDGE_API_KEY`를 켠 신뢰 LAN/VPN/tailnet/firewall/reverse proxy 뒤에서만 사용하십시오. 로그인 전 시작에는 `sudo loginctl enable-linger "$USER"`, 제거에는 `./uninstall.sh` 또는 `./uninstall.sh --purge-config`를 사용합니다.
+installer는 Linux user systemd용이며 CommandCode CLI 1.28.1 때문에 Node.js 22+를 요구합니다. 가능한 경우 CLI auth를 가져오고 private state는 `~/.config/commandcode-bridge`, 설치본은 `~/.local/share/commandcode-bridge`에 두며 안전한 기본값은 `127.0.0.1:9992`입니다. `0.0.0.0`은 `BRIDGE_API_KEY`를 켠 신뢰 LAN/VPN/tailnet/firewall/reverse proxy 뒤에서만 사용하십시오. 로그인 전 시작에는 `sudo loginctl enable-linger "$USER"`, 제거에는 `./uninstall.sh` 또는 `./uninstall.sh --purge-config`를 사용합니다.
 
 ```bash
 ./install.sh

README.md

@@ -8,7 +8,7 @@
 </p>
 
 <p align="center">
-  <a href="package.json"><img src="https://img.shields.io/badge/version-1.25.0.b-b57920?style=flat-square" alt="Version 1.25.0.b"></a>
+  <a href="package.json"><img src="https://img.shields.io/badge/version-1.28.1.a-b57920?style=flat-square" alt="Version 1.28.1.a"></a>
   <a href="src/model-catalog.ts"><img src="https://img.shields.io/badge/models-52-1f6f78?style=flat-square" alt="52 models"></a>
   <a href="package.json"><img src="https://img.shields.io/badge/Node.js-20%2B-9f4d2e?style=flat-square" alt="Node.js 20+"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-28231f?style=flat-square" alt="MIT License"></a>
@@ -20,7 +20,7 @@
 
 <!-- README-I18N:END -->
 
-CommandCode Bridge is a trusted-environment HTTP gateway for a CommandCode account. It presents standard OpenAI-compatible model and chat endpoints, routes work across eligible upstream credentials, and publishes an exact **55-model** catalog aligned with CommandCode **1.25.0**. The bridge version always tracks the current CommandCode CLI version with a letter suffix (for example **1.25.0.b**); the suffix marks bridge-only releases.
+CommandCode Bridge is a trusted-environment HTTP gateway for a CommandCode account. It presents standard OpenAI-compatible model and chat endpoints, routes work across eligible upstream credentials, and publishes an exact **56-model** catalog aligned with CommandCode **1.28.1**. The bridge version always tracks the current CommandCode CLI version with a letter suffix (for example **1.28.1.a**); the suffix marks bridge-only releases.
 
 [What it does](#what-it-does) · [Install](#install) · [Usage](#usage) · [How it works](#how-it-works) · [Repository layout](#repository-layout) · [Current limitations](#current-limitations) · [License](#license)
 
@@ -48,7 +48,7 @@ CommandCode Bridge is a trusted-environment HTTP gateway for a CommandCode accou
 
 ### Linux rootless installer
 
-The installer targets Linux user systemd, requires Node.js 22+ for CommandCode CLI 1.25.0, imports CLI auth when available, writes private state under `~/.config/commandcode-bridge`, installs under `~/.local/share/commandcode-bridge`, and safely defaults to `127.0.0.1:9992`. Use `0.0.0.0` only behind a trusted LAN/VPN/tailnet/firewall/reverse proxy with `BRIDGE_API_KEY`. Use `sudo loginctl enable-linger "$USER"` for pre-login startup; uninstall with `./uninstall.sh` or `./uninstall.sh --purge-config`.
+The installer targets Linux user systemd, requires Node.js 22+ for CommandCode CLI 1.28.1, imports CLI auth when available, writes private state under `~/.config/commandcode-bridge`, installs under `~/.local/share/commandcode-bridge`, and safely defaults to `127.0.0.1:9992`. Use `0.0.0.0` only behind a trusted LAN/VPN/tailnet/firewall/reverse proxy with `BRIDGE_API_KEY`. Use `sudo loginctl enable-linger "$USER"` for pre-login startup; uninstall with `./uninstall.sh` or `./uninstall.sh --purge-config`.
 
 ```bash
 ./install.sh

README.zh.md

@@ -8,7 +8,7 @@
 </p>
 
 <p align="center">
-  <a href="package.json"><img src="https://img.shields.io/badge/version-1.25.0.b-b57920?style=flat-square" alt="Version 1.25.0.b"></a>
+  <a href="package.json"><img src="https://img.shields.io/badge/version-1.28.1.a-b57920?style=flat-square" alt="Version 1.28.1.a"></a>
   <a href="src/model-catalog.ts"><img src="https://img.shields.io/badge/models-52-1f6f78?style=flat-square" alt="52 models"></a>
   <a href="package.json"><img src="https://img.shields.io/badge/Node.js-20%2B-9f4d2e?style=flat-square" alt="Node.js 20+"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-28231f?style=flat-square" alt="MIT License"></a>
@@ -20,7 +20,7 @@
 
 <!-- README-I18N:END -->
 
-CommandCode Bridge 是面向 CommandCode 账号的可信环境 HTTP 网关。它提供标准 OpenAI-compatible 模型与聊天端点，在符合条件的上游凭据之间路由请求，并发布与 CommandCode **1.25.0** 对齐的准确 **55-model** catalog。Bridge 版本始终跟随当前 CommandCode CLI 版本并在其后附加字母后缀（例如 **1.25.0.b**）；后缀表示仅限 Bridge 的发布。
+CommandCode Bridge 是面向 CommandCode 账号的可信环境 HTTP 网关。它提供标准 OpenAI-compatible 模型与聊天端点，在符合条件的上游凭据之间路由请求，并发布与 CommandCode **1.28.1** 对齐的准确 **56-model** catalog。Bridge 版本始终跟随当前 CommandCode CLI 版本并在其后附加字母后缀（例如 **1.28.1.a**）；后缀表示仅限 Bridge 的发布。
 
 [功能](#功能) · [安装](#安装) · [用法](#用法) · [工作原理](#工作原理) · [仓库布局](#仓库布局) · [当前限制](#当前限制) · [许可证](#许可证)
 
@@ -48,7 +48,7 @@ CommandCode Bridge 是面向 CommandCode 账号的可信环境 HTTP 网关。它
 
 ### Linux rootless installer
 
-installer 面向 Linux user systemd；由于 CommandCode CLI 1.25.0，需要 Node.js 22+。它会在可用时导入 CLI auth，把 private state 写入 `~/.config/commandcode-bridge`，安装到 `~/.local/share/commandcode-bridge`，安全默认值为 `127.0.0.1:9992`。只有在启用 `BRIDGE_API_KEY` 的可信 LAN/VPN/tailnet/firewall/reverse proxy 后才使用 `0.0.0.0`。登录前启动使用 `sudo loginctl enable-linger "$USER"`；卸载使用 `./uninstall.sh` 或 `./uninstall.sh --purge-config`。
+installer 面向 Linux user systemd；由于 CommandCode CLI 1.28.1，需要 Node.js 22+。它会在可用时导入 CLI auth，把 private state 写入 `~/.config/commandcode-bridge`，安装到 `~/.local/share/commandcode-bridge`，安全默认值为 `127.0.0.1:9992`。只有在启用 `BRIDGE_API_KEY` 的可信 LAN/VPN/tailnet/firewall/reverse proxy 后才使用 `0.0.0.0`。登录前启动使用 `sudo loginctl enable-linger "$USER"`；卸载使用 `./uninstall.sh` 或 `./uninstall.sh --purge-config`。
 
 ```bash
 ./install.sh

docs/PROCESS_LOG.md

@@ -86,12 +86,17 @@
 - Released bridge version `1.25.0.b`.
 - Added dashboard enable-all / disable-all controls beside the models heading.
 
-## Current status — 2026-08-15
+## 2026-08-19
+
+- Released bridge version `1.28.1.a` aligned with CommandCode CLI `1.28.1`.
+- Added `Qwen/Qwen3.8-27B` to the static catalog (56 models). GPT-5.6 Sol advertised price stays $5/$30 after the 1.27.1 standard-pricing note.
+
+## Current status — 2026-08-19
 
 - Branch: `main`, synchronized with `origin/main` when this status audit began.
-- Package: `commandcode-bridge` `1.25.0.b`, Node.js `>=20`, with `commandcode-bridge` and `commandcode-router` executables.
+- Package: `commandcode-bridge` `1.28.1.a`, Node.js `>=20`, with `commandcode-bridge` and `commandcode-router` executables.
 - API surface: authenticated OpenAI-compatible `/v1/models` and `/v1/chat/completions`, health endpoint, and same-origin dashboard configuration.
-- Model surface: 55 statically aligned models with live Provider API refresh when available.
+- Model surface: 56 statically aligned models with live Provider API refresh when available.
 - Routing surface: `daily_burn_priority`, `balance_priority`, `round_robin`, and `drain_first`, with per-key model scope, concurrency, cooldown, failover, and retry controls.
 - Deployment surface: Docker/Compose, Linux install/uninstall scripts, nginx and systemd release assets, and GitHub/GitLab CI definitions.
 - Verification: `npm run typecheck`, `npm run lint`, all 199 Vitest tests in 14 files, and `npm run build` pass on this workstation.

package.json

@@ -1,6 +1,6 @@
 {
   "name": "commandcode-bridge",
-  "version": "1.25.0.b",
+  "version": "1.28.1.a",
   "description": "OpenAI-compatible HTTP bridge for CommandCode's DeepSeek V4 Pro API path.",
   "type": "module",
   "license": "MIT",

src/config.ts

@@ -194,7 +194,7 @@ export function loadBridgeConfig(options: LoadBridgeConfigOptions = {}): BridgeC
     host: serverFromFile.host,
     port: serverFromFile.port,
     apiBase: (env.COMMANDCODE_API_BASE?.trim() || "https://api.commandcode.ai").replace(/\/+$/, ""),
-    cliVersion: env.COMMANDCODE_CLI_VERSION?.trim() || "1.25.0",
+    cliVersion: env.COMMANDCODE_CLI_VERSION?.trim() || "1.28.1",
     upstreamMode: parseUpstreamMode(env.COMMANDCODE_UPSTREAM_MODE),
     zdr: parseBoolean(env.COMMANDCODE_ZDR, false),
     defaultModel,

src/model-catalog.ts

@@ -314,6 +314,16 @@ export const COMMANDCODE_MODEL_DEFINITIONS: CommandCodeModelDefinition[] = [
     enabledByDefault: false,
     notes: "$2/M in · $6/M out",
   },
+  {
+    id: "Qwen/Qwen3.8-27B",
+    label: "Qwen 3.8 27B",
+    provider: "Qwen",
+    family: "qwen",
+    aliases: ["qwen3.8-27b", "Qwen3.8-27B", "alibaba/qwen3.8-27b"],
+    contextWindow: 262_000,
+    enabledByDefault: false,
+    notes: "$0.4/M in · $3/M out",
+  },
   {
     id: "Qwen/Qwen3.7-Max",
     label: "Qwen 3.7 Max",

src/openai.ts

@@ -66,6 +66,8 @@ export class CommandCodeEmptyVisibleResponseError extends Error {
 
 export interface StreamOpenAIChunksOptions extends CollectOpenAICompletionOptions {
   includeUsage?: boolean;
+  requestId?: string;
+  log?: { warn(payload: Record<string, unknown>, message: string): void };
 }
 
 export function mapUsageToOpenAI(usage: CommandCodeUsage | undefined): OpenAIUsage {
@@ -499,6 +501,20 @@ export async function* streamOpenAIChunks(
       }
     }
   } catch (error: unknown) {
+    options.log?.warn(
+      {
+        code: "commandcode_stream_error",
+        model: options.model,
+        request_id: options.requestId,
+        stream: true,
+        visible_content_length: visibleContentLength,
+        tool_call_count: toolCallCount,
+        saw_completion_signal: sawCompletionSignal,
+        finish_reason: finishReason ?? null,
+        error: error instanceof Error ? error.message : String(error),
+      },
+      "upstream stream failed mid-turn",
+    );
     yield sse(streamExceptionPayload(error));
     yield "data: [DONE]\n\n";
     return;

src/server.ts

@@ -16,6 +16,7 @@ import {
   publicModelList,
   publicModelObject,
   resolveModel,
+  type ResolvedModel,
 } from "./config.js";
 import {
   defaultModelCatalog,
@@ -654,13 +655,15 @@ export async function createApp(options: CreateAppOptions = {}): Promise<Fastify
   });
 
   app.post("/v1/chat/completions", async (request, reply) => {
+    let resolvedModel: ResolvedModel | undefined;
     try {
       const parsed = chatCompletionRequestSchema.parse(request.body);
       const openAIRequest = asOpenAIRequest(parsed);
-      const resolvedModel = resolveModel(openAIRequest.model, config);
+      resolvedModel = resolveModel(openAIRequest.model, config);
+      const resolved = resolvedModel;
       const useProviderChat =
         !eventPipelineAll &&
-        !isClaudeModelId(resolvedModel.upstreamModel) &&
+        !isClaudeModelId(resolved.upstreamModel) &&
         (config.upstreamMode === "provider" ||
           (config.upstreamMode === "auto" && providerAccessAvailable));
       const id = `chatcmpl_${randomUUID().replace(/-/g, "")}`;
@@ -704,6 +707,8 @@ export async function createApp(options: CreateAppOptions = {}): Promise<Fastify
             includeReasoning: config.includeReasoning,
             emptyVisibleResponsePolicy: config.emptyVisibleResponsePolicy,
             includeUsage: openAIRequest.stream_options?.include_usage ?? false,
+            requestId: request.id,
+            log: request.log,
           }),
         );
       }
@@ -766,18 +771,31 @@ export async function createApp(options: CreateAppOptions = {}): Promise<Fastify
             : error instanceof CommandCodeEventError
               ? error.upstreamBody
               : undefined;
+        const errorCode =
+          error instanceof CommandCodeHttpError
+            ? "commandcode_http_error"
+            : error instanceof CommandCodeEventError
+              ? "commandcode_event_error"
+              : error instanceof CommandCodeEmptyVisibleResponseError
+                ? "commandcode_empty_visible_response"
+                : "commandcode_empty_response";
+        if (!(error instanceof CommandCodeEmptyVisibleResponseError)) {
+          request.log.warn(
+            {
+              code: errorCode,
+              upstream_status: upstreamStatus,
+              request_id: request.id,
+              model: resolvedModel?.publicModel,
+              error: error.message,
+            },
+            "non-streaming chat failed with an upstream error",
+          );
+        }
         return reply.code(502).send({
           error: {
             message: error.message,
             type: "upstream_error",
-            code:
-              error instanceof CommandCodeHttpError
-                ? "commandcode_http_error"
-                : error instanceof CommandCodeEventError
-                  ? "commandcode_event_error"
-                  : error instanceof CommandCodeEmptyVisibleResponseError
-                    ? "commandcode_empty_visible_response"
-                    : "commandcode_empty_response",
+            code: errorCode,
             upstream_status: upstreamStatus,
             upstream_body: upstreamBody,
           },