billion-context-pi 0.1.49 → 0.1.50-pr.222.22

package/dist/index.js

@@ -3753,6 +3753,15 @@ function extractText(content) {
 function stripRefTag(text) {
   return text.replace(REF_TAG, "").replace(TRAILING_REF_TAG, "");
 }
+function countImageBlocks(content) {
+  if (!Array.isArray(content)) return 0;
+  let n = 0;
+  for (const block of content) {
+    const b = block;
+    if (b?.type === "image") n++;
+  }
+  return n;
+}
 function messageIdentity(message) {
   return JSON.stringify(normalizeIdentityValue(message, true));
 }
@@ -9172,12 +9181,29 @@ function collectCoveredMessageIds(state) {
   }
   return ids;
 }
-function estimateTokens(messages, coveredIds) {
+var IMAGE_TOKEN_COST = 1600;
+function modelSupportsImages(model) {
+  const input = model?.input;
+  return Array.isArray(input) && input.includes("image");
+}
+function collectImageTokens(entries, visionCapable) {
+  const out = /* @__PURE__ */ new Map();
+  if (!visionCapable) return out;
+  for (const e of entries) {
+    if (e.type !== "message") continue;
+    const n = countImageBlocks(e.message?.content);
+    if (n > 0) out.set(e.id, n * IMAGE_TOKEN_COST);
+  }
+  return out;
+}
+function estimateTokens(messages, coveredIds, imageTokensById) {
   let tokens = 0;
   for (const m of messages) {
     if (m.toolName === "compress") continue;
     if (coveredIds?.has(m.id)) continue;
     tokens += defaultCountTokens(m.text ?? "");
+    const img = imageTokensById?.get(m.id);
+    if (img) tokens += img;
   }
   return tokens;
 }
@@ -9306,12 +9332,13 @@ async function handleCompress(args, runtime, ctx, toolCallId) {
   if (typeof maybeRanges === "string") throw new Error(maybeRanges);
   const ranges = maybeRanges;
   if (ranges.length === 0) return "No ranges provided.";
-  const { state: initialState, coreMessages } = await runtime.stateFor(ctx);
+  const { state: initialState, coreMessages, entries } = await runtime.stateFor(ctx);
   const config = runtime.configFor(ctx);
   const modelId = ctx.model?.id ?? "default";
   const systemPromptText = getSystemPromptText(ctx);
   const systemPromptTokens = systemPromptText ? defaultCountTokens(systemPromptText) : 0;
-  const sentTokens = estimateTokens(coreMessages, collectCoveredMessageIds(initialState)) + systemPromptTokens;
+  const imageTokens = collectImageTokens(entries, modelSupportsImages(ctx.model));
+  const sentTokens = estimateTokens(coreMessages, collectCoveredMessageIds(initialState), imageTokens) + systemPromptTokens;
   const turn = runtime.core.processTurn({
     messages: coreMessages,
     state: initialState,
@@ -9321,7 +9348,7 @@ async function handleCompress(args, runtime, ctx, toolCallId) {
   const state = turn.state;
   const messages = turn.messages;
   const density = runtime.density.densityFor(modelId);
-  const beforeTokens = calibrateTokens(estimateTokens(messages, collectCoveredMessageIds(state)), density);
+  const beforeTokens = calibrateTokens(estimateTokens(messages, collectCoveredMessageIds(state), imageTokens), density);
   const summaryMaxChars = args.summaryMaxChars;
   const topLevelTopic = args.topic;
   debug.event("compress-in", {
@@ -9361,7 +9388,7 @@ async function handleCompress(args, runtime, ctx, toolCallId) {
     config,
     tokenCount: calibrateTokens(sentTokens, density)
   });
-  const afterTokens = calibrateTokens(estimateTokens(afterTurn.messages, collectCoveredMessageIds(applied.state)), density);
+  const afterTokens = calibrateTokens(estimateTokens(afterTurn.messages, collectCoveredMessageIds(applied.state), imageTokens), density);
   const reclaimed = Math.max(0, beforeTokens - afterTokens);
   const newBlocks = applied.state.blocks.slice(-blocksCreated);
   debug.event("compress-out", {
@@ -11038,13 +11065,13 @@ function makeStatusTool(runtime) {
   };
 }
 async function handleStatus(args, runtime, ctx) {
-  const { state, coreMessages } = await runtime.stateFor(ctx);
+  const { state, coreMessages, entries } = await runtime.stateFor(ctx);
   const config = runtime.configFor(ctx);
   const coveredIds = collectCoveredMessageIds(state);
   const modelId = ctx.model?.id ?? "default";
   const systemPromptText = getSystemPromptText(ctx);
   const systemPromptTokens = systemPromptText ? defaultCountTokens(systemPromptText) : 0;
-  const sentTokens = estimateTokens(coreMessages, coveredIds) + systemPromptTokens;
+  const sentTokens = estimateTokens(coreMessages, coveredIds, collectImageTokens(entries, modelSupportsImages(ctx.model))) + systemPromptTokens;
   const turn = runtime.core.processTurn({
     messages: coreMessages,
     state,
@@ -11528,12 +11555,14 @@ async function statusReport(runtime, ctx) {
   const realUsage = ctx.getContextUsage?.();
   const systemPromptText = getSystemPromptText(ctx);
   const systemPromptTokens = systemPromptText ? defaultCountTokens(systemPromptText) : 0;
-  const sessionTokens = realUsage?.tokens && realUsage.tokens > 0 ? realUsage.tokens : defaultCountTokens(coreMessages.map((m) => m.text ?? "").join("\n"));
+  const imageTokens = collectImageTokens(entries, modelSupportsImages(ctx.model));
+  const imageTokensTotal = [...imageTokens.values()].reduce((a, b) => a + b, 0);
+  const sessionTokens = realUsage?.tokens && realUsage.tokens > 0 ? realUsage.tokens : defaultCountTokens(coreMessages.map((m) => m.text ?? "").join("\n")) + imageTokensTotal;
   const coveredIds = collectCoveredMessageIds(state);
   const modelId = ctx.model?.id ?? "default";
-  const sentTokens = estimateTokens(coreMessages, coveredIds) + systemPromptTokens;
+  const sentTokens = estimateTokens(coreMessages, coveredIds, imageTokens) + systemPromptTokens;
   const turn = runtime.core.processTurn({ messages: coreMessages, state, config, tokenCount: calibrateTokens(sentTokens, runtime.density.densityFor(modelId)) });
-  const versionStr = "0.1.49" ? `billion-context-pi@${"0.1.49"}` : void 0;
+  const versionStr = "0.1.50" ? `billion-context-pi@${"0.1.50"}` : void 0;
   let text = buildStatusPanel({
     version: versionStr,
     tokenCount: sessionTokens,
@@ -11541,7 +11570,7 @@ async function statusReport(runtime, ctx) {
     state: turn.state,
     nudge: turn.nudge,
     modelContextLimit: config.modelContextLimit,
-    unprunedTokens: coreMessages.reduce((sum, m) => sum + defaultCountTokens(m.text ?? ""), 0),
+    unprunedTokens: coreMessages.reduce((sum, m) => sum + defaultCountTokens(m.text ?? "") + (imageTokens.get(m.id) ?? 0), 0),
     cacheUsages: cacheUsageSamples(entries ?? [])
   });
   const delegateUsage = getDelegateUsage();
@@ -11896,7 +11925,7 @@ async function autoInstallLatest(latest, extDirOverride) {
       "--no-fund",
       "--no-save"
     ];
-    const prevVersion = (await readPackageJson(join8(extDir, "package.json")))?.version ?? "0.1.49";
+    const prevVersion = (await readPackageJson(join8(extDir, "package.json")))?.version ?? "0.1.50";
     const { code, stderr } = await runNpmImpl(installArgs(latest), { cwd: npmDir, timeout: 6e4 });
     if (code !== 0) {
       logWarn("update", {
@@ -11909,7 +11938,7 @@ async function autoInstallLatest(latest, extDirOverride) {
     }
     const verify = await verifyInstall(npmDir, latest);
     if (!verify.ok) {
-      const rollbackTo = SEMVER_RE.test(prevVersion) ? prevVersion : "0.1.49";
+      const rollbackTo = SEMVER_RE.test(prevVersion) ? prevVersion : "0.1.50";
       logWarn("update", { event: "auto-install-verify-failed", latest, reason: verify.reason, rollbackTo });
       const rb = await runNpmImpl(installArgs(rollbackTo), { cwd: npmDir, timeout: 6e4 });
       logInfo("update", { event: "rollback", from: latest, to: rollbackTo, ok: rb.code === 0 });
@@ -11970,7 +11999,7 @@ async function checkForUpdate(autoUpdate, notify) {
     const runtimeVersion = await getRuntimeVersion();
     const latest = await fetchLatestVersion();
     if (!latest) return;
-    const current = runtimeVersion ?? "0.1.49";
+    const current = runtimeVersion ?? "0.1.50";
     const hasUpdate = isNewer(latest, current);
     debug.event("update-check", {
       current,
@@ -12049,7 +12078,7 @@ function wireSessionLifecycle(pi, runtime) {
     const sid = ctx.sessionManager.getSessionId();
     runtime.clearSessionTracking(sid);
     const modelInfo = ctx.model;
-    logInfo("session", { event: "start", sid, cwd: ctx.cwd, debug: runtime.adapter.debug ?? null, version: true ? "0.1.49" : null, model: modelInfo?.id ?? null, modelApi: modelInfo?.api ?? null, contextWindow: modelInfo?.contextWindow ?? null });
+    logInfo("session", { event: "start", sid, cwd: ctx.cwd, debug: runtime.adapter.debug ?? null, version: true ? "0.1.50" : null, model: modelInfo?.id ?? null, modelApi: modelInfo?.api ?? null, contextWindow: modelInfo?.contextWindow ?? null });
     try {
       await runtime.reloadConfig(ctx.cwd);
       setDelegateDisplayUsage(resolveDelegate(runtime.adapter).displayUsage);
@@ -12108,7 +12137,7 @@ function wireContextTransform(pi, runtime) {
       const realUsage = ctx.getContextUsage?.();
       const systemPromptText = getSystemPromptText(ctx);
       const systemPromptTokens = systemPromptText ? defaultCountTokens(systemPromptText) : 0;
-      const sentTokens = estimateTokens(coreMessages, coveredIds) + systemPromptTokens;
+      const sentTokens = estimateTokens(coreMessages, coveredIds, collectImageTokens(entries, modelSupportsImages(ctx.model))) + systemPromptTokens;
       let tokenCount = calibrateTokens(sentTokens, runtime.density.densityFor(modelId));
       if (ov.armed && config.modelContextLimit > 0) {
         ov.armed = false;