billion-context 0.1.82 → 0.1.83-pr.515.338

package/dist/index.js

@@ -48360,33 +48360,64 @@ async function contextFromRegistry(model, host) {
 function peekRegistryContext(model, host) {
   return registryLookup(cache2, model, host);
 }
+var VARIANT_SUFFIXES = [
+  "-interleaved-thinking",
+  "-thinking",
+  "-reasoning",
+  "-extended",
+  "-fast",
+  "-high",
+  "-medium",
+  "-low"
+];
+function modelVariants(name) {
+  const variants = [name];
+  let current = name;
+  for (; ; ) {
+    let stripped = false;
+    for (const suffix of VARIANT_SUFFIXES) {
+      if (current.length > suffix.length && current.endsWith(suffix)) {
+        current = current.slice(0, -suffix.length);
+        variants.push(current);
+        stripped = true;
+        break;
+      }
+    }
+    if (!stripped) break;
+  }
+  return variants;
+}
 function registryLookup(reg, model, host) {
   if (!reg || !model) return void 0;
   const provider = host ? providerFromHost(host) : void 0;
-  const candidates = provider ? [`${provider}/${model}`, model] : [model];
-  for (const key of candidates) {
-    const entry = reg[key];
-    const ctx = entry?.limit?.context;
-    if (typeof ctx === "number" && ctx > 0) return ctx;
-  }
-  if (provider === void 0) {
-    const suffix = `/${model}`;
-    let max;
-    const distinct = /* @__PURE__ */ new Set();
-    const parts = [];
-    for (const key of Object.keys(reg)) {
-      if (!key.endsWith(suffix)) continue;
-      const ctx = reg[key].limit?.context;
-      if (typeof ctx !== "number" || ctx <= 0) continue;
-      if (max === void 0 || ctx > max) max = ctx;
-      distinct.add(ctx);
-      parts.push(`${key}=${ctx}`);
-    }
-    if (max !== void 0 && distinct.size > 1 && !warnedConflicts.has(model)) {
-      warnedConflicts.add(model);
-      log("warn", `[acp-registry] conflicting context windows for "${model}" (${parts.join(", ")}) \u2014 using max ${max}`);
+  for (const name of modelVariants(model)) {
+    const candidates = provider ? [`${provider}/${name}`, name] : [name];
+    for (const key of candidates) {
+      const entry = reg[key];
+      const ctx = entry?.limit?.context;
+      if (typeof ctx === "number" && ctx > 0) return ctx;
+    }
+    if (provider === void 0) {
+      const suffix = `/${name}`;
+      let max;
+      const distinct = /* @__PURE__ */ new Set();
+      const parts = [];
+      for (const key of Object.keys(reg)) {
+        if (!key.endsWith(suffix)) continue;
+        const ctx = reg[key].limit?.context;
+        if (typeof ctx !== "number" || ctx <= 0) continue;
+        if (max === void 0 || ctx > max) max = ctx;
+        distinct.add(ctx);
+        parts.push(`${key}=${ctx}`);
+      }
+      if (max !== void 0) {
+        if (distinct.size > 1 && !warnedConflicts.has(name)) {
+          warnedConflicts.add(name);
+          log("warn", `[acp-registry] conflicting context windows for "${name}" (${parts.join(", ")}) \u2014 using max ${max}`);
+        }
+        return max;
+      }
     }
-    return max;
   }
   return void 0;
 }
@@ -54705,7 +54736,7 @@ function handlePluginStatus(conversationId2, res, fallbackLatest = false) {
   if (entry) entry.lastSeen = Date.now();
   const limit = session.metadata.effectiveContextLimit;
   const mem = remembered.get(session.id);
-  const modelContextLimit = typeof limit === "number" && limit > 0 ? limit : 2e5;
+  const modelContextLimit = typeof limit === "number" && limit > 0 ? limit : 0;
   let panel;
   try {
     panel = buildStatusPanel({
@@ -56729,6 +56760,9 @@ function logUnrecognizedPath(log2, url) {
     log2("info", `unrecognized path ${key}: forwarding unchanged; further occurrences suppressed`);
   }
 }
+function isModelDiscoveryPath(path18) {
+  return path18.replace(/\/+$/, "").endsWith("/models");
+}
 var BILI_HOP_HEADER = "x-bili-hop";
 function parseLauncherModelWindows(raw) {
   if (!raw) return {};
@@ -56998,6 +57032,13 @@ Content-Length: ${Buffer.byteLength(body)}\r
   }
   return server;
 }
+var SIDE_REQUEST_MAX_TOKENS = 200;
+function isSideRequest(parsed) {
+  if (!parsed || typeof parsed !== "object") return false;
+  const p2 = parsed;
+  const raw = p2.max_tokens ?? p2.max_completion_tokens ?? p2.max_output_tokens;
+  return typeof raw === "number" && raw > 0 && raw <= SIDE_REQUEST_MAX_TOKENS;
+}
 function isTrustedAdminOrigin(origin, host, trustedHosts) {
   if (!host || !trustedHosts.has(host.toLowerCase())) return false;
   if (!origin) return true;
@@ -57274,11 +57315,11 @@ ${bodyBuffer.toString("utf8")}`);
       const pluginWindow = pluginReportedContextWindow(req.headers);
       const launcherWindow = launcherContextWindow(model);
       const configuredWindow = resolveConfiguredContextLimit(opts.routes, embeddedUrl, model);
-      const peekWindow = host ? peekRegistryContext(model, host) : void 0;
+      const peekWindow = peekRegistryContext(model, host);
       let native = betaWindow ?? pluginWindow ?? launcherWindow ?? configuredWindow ?? peekWindow ?? lookupContextLimit(model);
       const operatorWindowTuned = resolveCompress(opts.routes, embeddedUrl, model, opts.compress).modelContextLimit !== void 0;
       nativeFromFallback = !betaWindow && !pluginWindow && !launcherWindow && !peekWindow && !configuredWindow && !operatorWindowTuned;
-      if (!native && host) {
+      if (!native) {
         native = await contextFromRegistry(model, host);
         if (native) nativeFromFallback = false;
       }
@@ -57379,7 +57420,6 @@ ${bodyBuffer.toString("utf8")}`);
     if (pluginAgent && !pluginConversation) pluginConversation = conversation;
     if (pluginAgent) {
       if (session.metadata.pluginAgent !== pluginAgent) session.metadata.pluginAgent = pluginAgent;
-      session.metadata.effectiveContextLimit = reqConfig.modelContextLimit;
       recordPluginSession(pluginConversation ?? conversation, session.id);
     }
     if (protocol === "responses" || protocol === "openai" || protocol === "anthropic") {
@@ -57389,6 +57429,21 @@ ${bodyBuffer.toString("utf8")}`);
       }
     }
     const pluginMode = pluginAgent !== void 0;
+    if (!countTokens && !responsesCompact && protocol !== null && isSideRequest(parsed)) {
+      log2("info", `[${session.id}] side request (max_tokens<=${SIDE_REQUEST_MAX_TOKENS}) \u2192 passthrough + tag strip only, kernel state untouched`);
+      const sidePrepared = {
+        body: bodyBuffer,
+        session,
+        processedMessages: [],
+        originalMessages: [],
+        protocol,
+        stream: parsed.stream === true,
+        compressInjected: false,
+        sidePassthrough: true
+      };
+      await forward(req, res, opts, bodyBuffer, sidePrepared, core, reqConfig, log2, route, instanceId, affinity);
+      return;
+    }
     const reqModel = parsed.model;
     const learnedMap = session.metadata.learnedContextLimits;
     const learnedLimit = (reqModel && learnedMap ? learnedMap[reqModel] : void 0) ?? session.metadata.learnedContextLimit;
@@ -57397,6 +57452,22 @@ ${bodyBuffer.toString("utf8")}`);
       reqConfig = { ...reqConfig, modelContextLimit: learnedLimit };
       nativeFromFallback = false;
       log2("info", `[${session.id}] self-healed context window: ${resolved} \u2192 ${learnedLimit} (learned from an upstream overflow)`);
+    } else if (nativeFromFallback && reqModel) {
+      const prevInput = session.stats.lastInputTokens ?? 0;
+      const prevWindow = session.metadata.lastTurnWindow;
+      const resolved = reqConfig.modelContextLimit;
+      if (prevWindow !== void 0 && prevInput > prevWindow && prevInput > resolved && prevInput >= 1e3) {
+        const map = session.metadata.learnedContextLimits ?? {};
+        const prev = map[reqModel];
+        if (prev === void 0 || prevInput > prev) {
+          map[reqModel] = prevInput;
+          session.metadata.learnedContextLimits = map;
+          markDirty(session);
+        }
+        reqConfig = { ...reqConfig, modelContextLimit: prevInput };
+        nativeFromFallback = false;
+        log2("info", `[${session.id}] self-healed context window upward: ${resolved} \u2192 ${prevInput} (a prior turn used ${prevInput} input tokens, exceeding the fallback window)`);
+      }
     }
     const nativeWindow = reqConfig.modelContextLimit;
     if (shouldReserveOutputHeadroom(protocol)) {
@@ -57410,6 +57481,8 @@ ${bodyBuffer.toString("utf8")}`);
       }
       if (reserved !== reqConfig.modelContextLimit) reqConfig = { ...reqConfig, modelContextLimit: reserved };
     }
+    session.metadata.effectiveContextLimit = reqConfig.modelContextLimit;
+    session.metadata.lastTurnWindow = reqConfig.modelContextLimit;
     acquireInFlight(session);
     try {
       await withSessionLock(session, async () => {
@@ -57474,7 +57547,7 @@ ${bodyBuffer.toString("utf8")}`);
     }
   }
   if (!prepared) {
-    if (protocol === null && !opts.passthrough) {
+    if (protocol === null && !opts.passthrough && !isModelDiscoveryPath(urlPath)) {
       logUnrecognizedPath(log2, req.url ?? "");
     }
     await forward(req, res, opts, bodyBuffer, null, core, reqConfig, log2, route, instanceId, void 0);
@@ -58311,7 +58384,7 @@ ${hdrText}
   }
   let responseBody = upstream.body;
   try {
-    if (prepared !== null && prepared.stream && maxFakeCompletionRetries() > 0) {
+    if (prepared !== null && prepared.stream && !prepared.sidePassthrough && maxFakeCompletionRetries() > 0) {
       const resolvedBuf = await resolveFakeCompletion(upstream.body, {
         protocol: prepared.protocol,
         body,