refactor(log): unify request lifecycle logging with session-colored tags

Collapse scattered per-request console lines (request/routing/auth/pending/
usage/stream-usage/stream) into 3 correlated lines: request, transform,
done. Add stable per-session color tag so concurrent request lines are
easy to follow, surface thinking intent, always-on full error logging
for debug, re-enable warn level, and uppercase keyword labels. Also fix
usage overview cards wrapping (5 cards -> grid-cols-5).

Co-authored-by: Cursor <cursoragent@cursor.com>

src/sse/handlers/chat.js

@@ -48,15 +48,9 @@ export async function handleChat(request, clientRawRequest = null) {
   }
   cacheClaudeHeaders(clientRawRequest.headers);
 
-  // Log request endpoint and model
-  const url = new URL(request.url);
   const modelStr = body.model;
 
-  // Count messages (support both messages[] and input[] formats)
-  const msgCount = body.messages?.length || body.input?.length || 0;
-  const toolCount = body.tools?.length || 0;
-  const effort = body.reasoning_effort || body.reasoning?.effort || null;
-  log.request("POST", `${url.pathname} | ${modelStr} | ${msgCount} msgs${toolCount ? ` | ${toolCount} tools` : ""}${effort ? ` | effort=${effort}` : ""}`);
+  // Request summary is emitted as the unified "▶" line in chatCore (has fmt/thinking/account)
 
   // Log API key (masked)
   const authHeader = request.headers.get("Authorization");
@@ -191,12 +185,7 @@ async function handleSingleModelChat(body, modelStr, clientRawRequest = null, re
 
   const { provider, model } = modelInfo;
 
-  // Log model routing (alias → actual model)
-  if (modelStr !== `${provider}/${model}`) {
-    log.info("ROUTING", `${modelStr} → ${provider}/${model}`);
-  } else {
-    log.info("ROUTING", `Provider: ${provider}, Model: ${model}`);
-  }
+  // Routing shown in the unified "▶" line (client model → provider/model)
 
   // Extract userAgent from request
   const userAgent = request?.headers?.get("user-agent") || "";
@@ -225,9 +214,7 @@ async function handleSingleModelChat(body, modelStr, clientRawRequest = null, re
       return errorResponse(lastStatus || HTTP_STATUS.SERVICE_UNAVAILABLE, lastError || "All accounts unavailable");
     }
 
-    // Log account selection
-    log.info("AUTH", `\x1b[32mUsing ${provider} account: ${credentials.connectionName}\x1b[0m`);
-
+    // Account selection shown in the unified "▶" line (acc:...)
     const refreshedCredentials = await checkAndRefreshToken(provider, credentials);
 
     // Ensure real project ID is available for providers that need it (P0 fix: cold miss)
@@ -288,7 +275,7 @@ async function handleSingleModelChat(body, modelStr, clientRawRequest = null, re
     const { shouldFallback } = await markAccountUnavailable(credentials.connectionId, result.status, result.error, provider, model, result.resetsAtMs);
 
     if (shouldFallback) {
-      log.warn("AUTH", `Account ${credentials.connectionName} unavailable (${result.status}), trying fallback`);
+      log.warn("FALLBACK", `⇄ ACC:${credentials.connectionName} UNAVAILABLE (${result.status}) → NEXT ACCOUNT`);
       excludeConnectionIds.add(credentials.connectionId);
       lastError = result.error;
       lastStatus = result.status;

src/sse/utils/logger.js

@@ -13,6 +13,49 @@ function formatTime() {
   return new Date().toLocaleTimeString("en-US", { hour12: false });
 }
 
+// Colored-dot tags to correlate request lines by session (same session → same color)
+const REQ_TAGS = ["🟢", "🔵", "🟣", "🟡", "🟠", "🔴", "⚪", "🟤"];
+let tagCursor = 0;
+
+// Allocate next rotating tag (fallback when no session seed available)
+export function nextTag() {
+  const tag = REQ_TAGS[tagCursor % REQ_TAGS.length];
+  tagCursor++;
+  return tag;
+}
+
+// Stable tag derived from a session/connection seed: same seed always maps to the same color
+export function tagForSession(seed) {
+  if (!seed) return nextTag();
+  let h = 0;
+  for (let i = 0; i < seed.length; i++) h = (h * 31 + seed.charCodeAt(i)) | 0;
+  return REQ_TAGS[Math.abs(h) % REQ_TAGS.length];
+}
+
+// Print one correlated line: [time] tag symbol message
+export function line(tag, symbol, message) {
+  if (LEVEL > LOG_LEVELS.INFO) return;
+  console.log(`[${formatTime()}] ${tag} ${symbol} ${message}`);
+}
+
+// Like line() but always printed regardless of LOG_LEVEL (errors must never be hidden)
+export function errorLine(tag, symbol, message) {
+  console.log(`[${formatTime()}] ${tag} ${symbol} ${message}`);
+}
+
+// Format thinking intent for the request line ("high(10k)" / "off" / "auto")
+export function fmtThink(intent) {
+  if (!intent || !intent.mode) return null;
+  if (intent.mode === "none") return "off";
+  if (intent.mode === "auto") return "auto";
+  if (intent.mode === "budget") {
+    const k = intent.budget >= 1000 ? `${Math.round(intent.budget / 1000)}k` : `${intent.budget}`;
+    return k;
+  }
+  if (intent.mode === "level") return intent.level;
+  return null;
+}
+
 function formatData(data) {
   if (!data) return "";
   if (typeof data === "string") return data;
@@ -40,7 +83,7 @@ export function info(tag, message, data) {
 export function warn(tag, message, data) {
   if (LEVEL <= LOG_LEVELS.WARN) {
     const dataStr = data ? ` ${formatData(data)}` : "";
-    // console.warn(`[${formatTime()}] ⚠️  [${tag}] ${message}${dataStr}`);
+    console.warn(`[${formatTime()}] ⚠️  [${tag}] ${message}${dataStr}`);
   }
 }