@youdie006/prodex 0.21.0 → 0.21.2

package/dist/chatgpt-browser.js

@@ -1872,6 +1872,7 @@ export async function sendChatGptPrompt(options) {
             emitProgress("selecting", `attached ${uploaded.attached.join(", ")}`);
         }
         const toolLabels = (options.tools ?? []).map(resolveComposerToolLabel);
+        const wantsDeepResearch = toolLabels.includes(DEEP_RESEARCH_TOOL_LABEL);
         if (toolLabels.length > 0)
             emitProgress("selecting", `tools=${toolLabels.join(", ")}`);
         await insertComposerTextViaCdp(cdp, options.prompt, page, toolLabels);
@@ -1915,6 +1916,26 @@ export async function sendChatGptPrompt(options) {
             }
         }
         dbgSend(`submit posted=${promptPosted} submitButtonFound=${submitButtonFound}`);
+        // Deep research does not begin when the prompt posts: it shows a start
+        // control with a countdown ring. Press it instead of trusting the timer -
+        // a run left waiting sat with zero assistant messages for 30+ minutes.
+        if (wantsDeepResearch) {
+            const startDeadline = Date.now() + 60_000;
+            let pressed = false;
+            while (Date.now() < startDeadline) {
+                const start = await cdp.evaluate(deepResearchStartButtonRectExpression());
+                if (start.ok && start.x !== undefined && start.y !== undefined) {
+                    await dispatchMouseClickAt(cdp, start.x, start.y);
+                    pressed = true;
+                    emitProgress("selecting", "deep research started");
+                    break;
+                }
+                await sleep(1_000);
+            }
+            if (!pressed) {
+                sendWarnings.push("deep_research_start_not_found: no start control appeared for the deep research run. If ChatGPT asked a clarifying question instead, answer it with a follow-up consult in the same thread.");
+            }
+        }
     }
     finally {
         cdp.close();
@@ -1977,13 +1998,38 @@ export async function sendChatGptPrompt(options) {
         }
         throw acceptanceTimeoutError({ timeoutMs, composerStillHasText, submitButtonFound });
     }
+    // Pin the conversation the prompt actually landed in. The browser is shared
+    // (other agents, the user, tooling), and a tab that moves mid-wait made
+    // prodex read a DIFFERENT conversation and save it as this consult's answer -
+    // silently, with a receipt (caught live). Nothing about that is recoverable
+    // after the fact, so the wait either stays on this thread or fails loudly.
+    const pinnedThreadUrl = finalState?.url;
+    let recoveredNavigations = 0;
     const answerIsStable = createChatGptAnswerStabilityTracker();
     while (Date.now() - started < timeoutMs) {
         await sleep(1000);
         try {
             finalState = await evaluateOnPage(page, answerExpression());
+            if (pinnedThreadUrl && finalState?.url && !chatGptUrlsReferToSameTarget(finalState.url, pinnedThreadUrl)) {
+                if (recoveredNavigations >= 2) {
+                    throw new ChatGptBrowserBlockerError({
+                        code: "thread_navigated_away",
+                        message: "The browser tab was moved to a different ChatGPT conversation while this consult was waiting for its answer.",
+                        retryable: true,
+                        next_step: `Keep the dedicated browser on the consult thread, then fetch the answer with \`prodex pro browser recover --target-url ${pinnedThreadUrl}\`.`,
+                        thread: pinnedThreadUrl
+                    });
+                }
+                recoveredNavigations += 1;
+                sendWarnings.push(`thread_navigated_away_recovered: something moved the tab to another conversation mid-wait; prodex navigated back to ${pinnedThreadUrl}.`);
+                await evaluateOnPage(page, `location.assign(${JSON.stringify(pinnedThreadUrl)})`);
+                await sleep(3_000);
+                continue;
+            }
         }
-        catch {
+        catch (error) {
+            if (error instanceof ChatGptBrowserBlockerError)
+                throw error;
             // Transient CDP failure while the answer is streaming: retry. A throw here
             // would discard an already-streamed partial answer and skip the salvage
             // path below, so keep the last good state and poll again until timeout.
@@ -2564,10 +2610,31 @@ export async function attachFilesToComposer(cdp, absolutePaths, options = {}) {
 // check has to ignore the token or it reads as leftover contamination.
 // ---------------------------------------------------------------------------
 const COMPOSER_TOOL_ALIASES = [
-    { label: "Deep research", aliases: ["deep research", "deep-research", "deepresearch", "deep", "research"] },
-    { label: "Web search", aliases: ["web search", "web-search", "websearch", "search", "web"] },
-    { label: "Create image", aliases: ["create image", "create-image", "image", "img"] }
+    // menuText carries every string the menu row is known to render. ChatGPT
+    // dropped the visible "Deep research" title at one point and left only its
+    // description, which made a title-only lookup fail (measured live), so each
+    // tool is found by title OR description.
+    {
+        label: "Deep research",
+        aliases: ["deep research", "deep-research", "deepresearch", "deep", "research"],
+        menuText: ["Deep research", "Get a detailed report"]
+    },
+    {
+        label: "Web search",
+        aliases: ["web search", "web-search", "websearch", "search", "web"],
+        menuText: ["Web search", "Find real-time news and info"]
+    },
+    {
+        label: "Create image",
+        aliases: ["create image", "create-image", "image", "img"],
+        menuText: ["Create image", "Visualize anything"]
+    }
 ];
+/** Every string the tools menu may render for this tool. */
+export function composerToolMenuTexts(label) {
+    const known = COMPOSER_TOOL_ALIASES.find((tool) => tool.label === label);
+    return known ? [...known.menuText] : [label];
+}
 /**
  * Map what a caller typed to the label ChatGPT renders. An unknown value is
  * passed through unchanged, so a tool ChatGPT adds tomorrow is reachable by
@@ -2594,11 +2661,11 @@ export function composerToolsButtonRectExpression() {
 }
 /** Click point for a tools-menu entry, matched by its visible label. */
 export function composerToolEntryRectExpression(label) {
-    const labelJson = JSON.stringify(label);
+    const candidatesJson = JSON.stringify(composerToolMenuTexts(label).map((text) => text.toLowerCase()));
     return `(() => {${CLICK_POINT_SNIPPET}
-    const wanted = ${labelJson}.trim().toLowerCase();
+    const candidates = ${candidatesJson};
     const leaves = [...document.querySelectorAll("div,span,button,a")].filter((el) => el.children.length === 0);
-    const leaf = leaves.find((el) => (el.textContent || "").trim().toLowerCase() === wanted);
+    const leaf = leaves.find((el) => candidates.includes((el.textContent || "").trim().toLowerCase()));
     if (!leaf) {
       const available = [...new Set(leaves.map((el) => (el.textContent || "").trim()).filter((t) => t.length > 1 && t.length < 30))].slice(0, 20);
       return { ok: false, reason: "tool not found in the composer tools menu", available };
@@ -2611,8 +2678,12 @@ export function composerToolEntryRectExpression(label) {
 export function activeComposerToolsExpression(labels) {
     const labelsJson = JSON.stringify(labels);
     return `(() => {
+    // The token normally lands inside the editor, but some builds render it as
+    // a pill in the surrounding composer form - check both before deciding a
+    // selection did not take.
     const el = document.querySelector('#prompt-textarea,[contenteditable="true"]');
-    const text = el ? (el.innerText || "") : "";
+    const form = el ? (el.closest("form") || el.parentElement) : null;
+    const text = (el ? el.innerText || "" : "") + String.fromCharCode(10) + (form ? form.innerText || "" : "");
     return { ok: true, active: ${labelsJson}.filter((label) => text.includes(label)) };
   })()`;
 }
@@ -2649,8 +2720,9 @@ export function composerTextStateExpression(expectedText, toolLabels = []) {
 export async function enableComposerTools(cdp, labels) {
     const enabled = [];
     for (const label of labels) {
-        const already = await cdp.evaluate(activeComposerToolsExpression([label]));
-        if ((already?.active ?? []).includes(label)) {
+        const activeProbe = composerToolMenuTexts(label);
+        const already = await cdp.evaluate(activeComposerToolsExpression(activeProbe));
+        if ((already?.active ?? []).length > 0) {
             enabled.push(label);
             continue;
         }
@@ -2678,12 +2750,33 @@ export async function enableComposerTools(cdp, labels) {
         const activeDeadline = Date.now() + 8_000;
         let active = false;
         for (;;) {
-            const state = await cdp.evaluate(activeComposerToolsExpression([label]));
-            active = (state?.active ?? []).includes(label);
+            const state = await cdp.evaluate(activeComposerToolsExpression(activeProbe));
+            active = (state?.active ?? []).length > 0;
             if (active || Date.now() >= activeDeadline)
                 break;
             await sleep(250);
         }
+        if (!active) {
+            // One retry: the menu can close on a click that lands as the popover is
+            // still settling, which looks identical to a refused selection.
+            const retryButton = await cdp.evaluate(composerToolsButtonRectExpression());
+            if (retryButton.ok && retryButton.x !== undefined && retryButton.y !== undefined) {
+                await dispatchMouseClickAt(cdp, retryButton.x, retryButton.y);
+                await sleep(1_000);
+                const retryEntry = await cdp.evaluate(composerToolEntryRectExpression(label));
+                if (retryEntry.ok && retryEntry.x !== undefined && retryEntry.y !== undefined) {
+                    await dispatchMouseClickAt(cdp, retryEntry.x, retryEntry.y);
+                    const retryDeadline = Date.now() + 8_000;
+                    for (;;) {
+                        const state = await cdp.evaluate(activeComposerToolsExpression(activeProbe));
+                        active = (state?.active ?? []).length > 0;
+                        if (active || Date.now() >= retryDeadline)
+                            break;
+                        await sleep(250);
+                    }
+                }
+            }
+        }
         if (!active)
             throw new Error(`Selected "${label}" but the composer never showed it as active.`);
         enabled.push(label);
@@ -2778,6 +2871,27 @@ export function chunkComposerText(text, size = COMPOSER_INSERT_CHUNK_CHARS) {
         chunks.push(current);
     return chunks;
 }
+/**
+ * The start control a deep research run shows after the prompt posts. It has a
+ * countdown ring and starts on its own eventually, but prodex must not depend
+ * on that: a run left waiting produced zero assistant messages for 30+ minutes
+ * (measured live), because nothing pressed it.
+ */
+export function deepResearchStartButtonRectExpression() {
+    return `(() => {${CLICK_POINT_SNIPPET}
+    const buttons = [...document.querySelectorAll('button,[role="button"]')];
+    const target = buttons.find((b) => {
+      const text = ((b.innerText || "") + " " + (b.getAttribute("aria-label") || "")).trim();
+      // "Start dictation" and "Start Voice" live in the same composer and
+      // would otherwise swallow this click - pressing the microphone instead
+      // of starting the research (measured live).
+      if (/dictation|voice|mic|음성|받아쓰기/i.test(text)) return false;
+      return /^(start|시작)\\s*$|start research|시작하기|리서치 시작|조사 시작/i.test(text);
+    });
+    if (!target) return { ok: false, reason: "no deep research start button" };
+    return clickPoint(target);
+  })()`;
+}
 export function submitExpression() {
     return `(() => {
     ${composerExpressionHelpers()}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@youdie006/prodex",
-  "version": "0.21.0",
+  "version": "0.21.2",
   "description": "Local receipt bus for coordinating Codex execution with ChatGPT Pro/Projects consultation.",
   "author": "youdie006",
   "license": "MIT",