@timo972/cc-router 0.10.0 → 0.10.1

package/CHANGELOG.md

@@ -8,6 +8,32 @@ This project follows [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ## [Unreleased]
 
+---
+
+## [0.10.1] — 2026-08-20
+
+### Fixed
+
+- A passed-through upstream 5xx on the Anthropic path is logged and counted.
+  The proxy is byte-transparent and only special-cased 401/429/529, so a
+  plain upstream 500 left no trace: an overnight Anthropic 500 stopped an
+  unattended Claude session while the daemon log showed nothing and the
+  stats reported a clean night. It now produces an `[ERROR]` line, an
+  activity entry (`upstream-error`), and error counts — with no cooldown,
+  since a plain 5xx says nothing about the account's capacity and can even
+  be request-specific.
+- A client-cancelled stream abort is no longer logged as an error on the
+  OpenAI path. The Codex CLI aborts streams routinely, and each abort
+  rejects the relay's body read — the log printed an `[ERROR] ... relay
+  failed` line for every one (eight hours of them in one overnight
+  session) while the stats correctly classified them as cancellations. The
+  log line now waits for the cancellation check, so it fires only for real
+  relay failures.
+
+---
+
+## [0.10.0] — 2026-08-19
+
 ### Added
 
 - OpenAI/Codex account usage is fetched proactively, so `cc-router status`
@@ -52,6 +78,29 @@ This project follows [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
   one row when the selection steps past its bottom or top edge, stays put while
   the selection moves inside it, and re-clamps when new entries push the
   selected row (which is timestamp-anchored) out of the stored window.
+- OpenAI/Codex sticky session routing: sessions pin to one account for prompt-cache
+  locality (`session_id` → `x-claude-code-session-id` → `prompt_cache_key`), with
+  load- and headroom-aware selection for new sessions.
+- Codex usage tracking from `x-codex-*` response headers: default 5h/weekly windows
+  plus dynamically discovered model-scoped metered buckets, credits, and plan.
+- Scoped cooldowns on upstream failures: bucket-scoped via `x-codex-active-limit`,
+  account-global otherwise; local 429/503 responses when no account is eligible.
+- Dashboard: OpenAI accounts now show 5h/weekly bars, per-bucket rows, credits,
+  plan, request/error/in-flight/session counts, and cooldown state.
+- Unprefixed `gpt-*` models route to OpenAI. The Codex CLI writes the bare slug
+  from its own registry — `model = "gpt-5.6-sol"` in `config.toml`, or whatever
+  its `/model` picker selects — and an unprefixed name went to the Claude path,
+  where `/v1/responses` answers `501 Not Implemented`. No configuration could
+  redirect it, because `openAIAliases` is only consulted for names that are
+  already prefixed; those aliases now apply to the bare form as well. Every
+  other unprefixed model still routes to Claude.
+
+### Changed
+
+- OpenAI account records persist `scopes`, `sessionLimitPercent`, and
+  `weeklyLimitPercent`.
+- The stateless OpenAI round-robin picker was removed in favor of
+  `OpenAITokenPool`.
 
 ### Fixed
 
@@ -143,39 +192,6 @@ This project follows [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
   on `CC_ROUTER_DAEMON`, which the LaunchAgent and systemd unit never set —
   they set `CC_ROUTER_SERVICE` — so every service-managed instance left no PID
   behind and took the weaker port-based stop path.
-
----
-
-## [0.10.0] — 2026-08-18
-
-### Added
-
-- OpenAI/Codex sticky session routing: sessions pin to one account for prompt-cache
-  locality (`session_id` → `x-claude-code-session-id` → `prompt_cache_key`), with
-  load- and headroom-aware selection for new sessions.
-- Codex usage tracking from `x-codex-*` response headers: default 5h/weekly windows
-  plus dynamically discovered model-scoped metered buckets, credits, and plan.
-- Scoped cooldowns on upstream failures: bucket-scoped via `x-codex-active-limit`,
-  account-global otherwise; local 429/503 responses when no account is eligible.
-- Dashboard: OpenAI accounts now show 5h/weekly bars, per-bucket rows, credits,
-  plan, request/error/in-flight/session counts, and cooldown state.
-- Unprefixed `gpt-*` models route to OpenAI. The Codex CLI writes the bare slug
-  from its own registry — `model = "gpt-5.6-sol"` in `config.toml`, or whatever
-  its `/model` picker selects — and an unprefixed name went to the Claude path,
-  where `/v1/responses` answers `501 Not Implemented`. No configuration could
-  redirect it, because `openAIAliases` is only consulted for names that are
-  already prefixed; those aliases now apply to the bare form as well. Every
-  other unprefixed model still routes to Claude.
-
-### Changed
-
-- OpenAI account records persist `scopes`, `sessionLimitPercent`, and
-  `weeklyLimitPercent`.
-- The stateless OpenAI round-robin picker was removed in favor of
-  `OpenAITokenPool`.
-
-### Fixed
-
 - A refresh token the OAuth server rejects as terminally expired
   (`400 invalid_grant`) is no longer retried forever. Every rejection was
   treated as transient, so the five-minute refresh loop kept re-POSTing a token

package/dist/proxy/openai-ingress.js

@@ -304,6 +304,7 @@ export async function runOpenAIIngress(opts) {
     };
     let finalStatus = upstream.status;
     let relayFailed = false;
+    let relayFailureMessage = "";
     const relayReport = { upstreamReportedFailure: false };
     try {
         const result = await relay(upstream, res, entry, relayReport);
@@ -316,7 +317,7 @@ export async function runOpenAIIngress(opts) {
         // can do is tear the connection down.
         relayFailed = true;
         const message = error instanceof Error ? error.message : String(error);
-        logError(account.id, 502, `openai response relay failed: ${message}`);
+        relayFailureMessage = message;
         // The recorded status is what this request *became*, which is a failure
         // whether or not another HTTP response can still be sent. Leaving it at
         // the upstream's 200 in the headers-already-sent case produced an
@@ -346,6 +347,15 @@ export async function runOpenAIIngress(opts) {
     // upstream announce a failure. Only the truncation is the disconnect's to
     // explain away.
     const clientCancelled = clientGone.signal.aborted && !relayReport.upstreamReportedFailure;
+    // Logged only now, after the cancellation classification: the Codex CLI
+    // aborts streams routinely (a superseded turn, Ctrl-C, a pane closing),
+    // and each abort rejects the relay's body read with "This operation was
+    // aborted". Logging that from the catch block printed an [ERROR] line for
+    // every one — eight hours of them in one unattended overnight session —
+    // while the stats correctly ignored them. The log now matches the stats.
+    if (relayFailed && !clientCancelled) {
+        logError(account.id, 502, `openai response relay failed: ${relayFailureMessage}`);
+    }
     // Activity/stats must reflect what the client actually received, not just
     // the raw upstream signal: the non-streaming collector can synthesize a
     // local 502 from an upstream 200 whose SSE stream ended in

package/dist/proxy/server.js

@@ -1037,6 +1037,25 @@ export async function startServer(opts = {}) {
                         : "service-overloaded";
                     logError(account.id, 529, "Service overloaded — cooldown 30s");
                 }
+                else if (status >= 500) {
+                    // Any other upstream 5xx passes through byte-transparent — but it
+                    // must not pass through the DIAGNOSTICS silently: an overnight
+                    // Anthropic 500 stopped an unattended Claude session while this
+                    // log showed nothing and the stats reported a clean night. The
+                    // question "did the proxy or the upstream fail?" was only
+                    // answerable by cross-referencing the client's own transcript.
+                    // Unlike 429/529 this takes no cooldown: a plain 5xx says nothing
+                    // about the account's capacity and can even be request-specific,
+                    // so cooling the account down would punish it for upstream's (or
+                    // the request's) problem.
+                    stats.totalErrors++;
+                    account.errorCount++;
+                    pendingLog.type = "error";
+                    pendingLog.details = route
+                        ? routeFailureDetails(route, "upstream-error")
+                        : "upstream-error";
+                    logError(account.id, status, "Upstream server error (passed through)");
+                }
                 // ── Capture rate limit utilization from response headers ────────────
                 applyRateLimitHeaders(account, proxyRes.headers);
                 const entry = pendingLog;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@timo972/cc-router",
-  "version": "0.10.0",
+  "version": "0.10.1",
   "description": "Cache-aware session router for Claude Max OAuth tokens — use multiple Claude Max accounts with Claude Code",
   "type": "module",
   "bin": {