@steipete/oracle 0.6.1 → 0.7.1

package/README.md

@@ -21,26 +21,29 @@ Use `npx -y @steipete/oracle …` (not `pnpx`)—pnpx's sandboxed cache can’t
 
 ```bash
 # Copy the bundle and paste into ChatGPT
-npx @steipete/oracle --render --copy -p "Review the TS data layer for schema drift" --file "src/**/*.ts,*/*.test.ts"
+npx -y @steipete/oracle --render --copy -p "Review the TS data layer for schema drift" --file "src/**/*.ts,*/*.test.ts"
 
 # Minimal API run (expects OPENAI_API_KEY in your env)
-npx @steipete/oracle -p "Write a concise architecture note for the storage adapters" --file src/storage/README.md
+npx -y @steipete/oracle -p "Write a concise architecture note for the storage adapters" --file src/storage/README.md
 
 # Multi-model API run
-npx @steipete/oracle -p "Cross-check the data layer assumptions" --models gpt-5.1-pro,gemini-3-pro --file "src/**/*.ts"
+npx -y @steipete/oracle -p "Cross-check the data layer assumptions" --models gpt-5.1-pro,gemini-3-pro --file "src/**/*.ts"
 
 # Preview without spending tokens
-npx @steipete/oracle --dry-run summary -p "Check release notes" --file docs/release-notes.md
+npx -y @steipete/oracle --dry-run summary -p "Check release notes" --file docs/release-notes.md
 
 # Browser run (no API key, will open ChatGPT)
-npx @steipete/oracle --engine browser -p "Walk through the UI smoke test" --file "src/**/*.ts"
+npx -y @steipete/oracle --engine browser -p "Walk through the UI smoke test" --file "src/**/*.ts"
+
+# Gemini browser mode (no API key; uses Chrome cookies from gemini.google.com)
+npx -y @steipete/oracle --engine browser --model gemini-3-pro --prompt "a cute robot holding a banana" --generate-image out.jpg --aspect 1:1
 
 # Sessions (list and replay)
-npx @steipete/oracle status --hours 72
-npx @steipete/oracle session <id> --render
+npx -y @steipete/oracle status --hours 72
+npx -y @steipete/oracle session <id> --render
 
 # TUI (interactive, only for humans)
-npx @steipete/oracle tui
+npx -y @steipete/oracle tui
 ```
 
 Engine auto-picks API when `OPENAI_API_KEY` is set, otherwise browser; browser is stable on macOS and works on Linux and Windows. On Linux pass `--browser-chrome-path/--browser-cookie-path` if detection fails; on Windows prefer `--browser-manual-login` or inline cookies if decryption is blocked.
@@ -49,6 +52,8 @@ Engine auto-picks API when `OPENAI_API_KEY` is set, otherwise browser; browser i
 
 **CLI**
 - API mode expects API keys in your environment: `OPENAI_API_KEY` (GPT-5.x), `GEMINI_API_KEY` (Gemini 3 Pro), `ANTHROPIC_API_KEY` (Claude Sonnet 4.5 / Opus 4.1).
+- Gemini browser mode uses Chrome cookies instead of an API key—just be logged into `gemini.google.com` in Chrome (no Python/venv required).
+- If your Gemini account can’t access “Pro”, Oracle auto-falls back to a supported model for web runs (and logs the fallback in verbose mode).
 - Prefer API mode or `--copy` + manual paste; browser automation is experimental.
 - Browser support: stable on macOS; works on Linux (add `--browser-chrome-path/--browser-cookie-path` when needed) and Windows (manual-login or inline cookies recommended when app-bound cookies block decryption).
 - Remote browser service: `oracle serve` on a signed-in host; clients use `--remote-host/--remote-token`.
@@ -109,6 +114,9 @@ npx -y @steipete/oracle oracle-mcp
 | `--dry-run [summary\|json\|full]` | Preview without sending. |
 | `--remote-host`, `--remote-token` | Use a remote `oracle serve` host (browser). |
 | `--remote-chrome <host:port>` | Attach to an existing remote Chrome session (browser). |
+| `--youtube <url>` | YouTube video URL to analyze (Gemini browser mode). |
+| `--generate-image <file>` | Generate image and save to file (Gemini browser mode). |
+| `--edit-image <file>` | Edit existing image with `--output` (Gemini browser mode). |
 | `--azure-endpoint`, `--azure-deployment`, `--azure-api-version` | Target Azure OpenAI endpoints (picks Azure client automatically). |
 
 ## Configuration

package/dist/bin/oracle-cli.js

@@ -18,6 +18,7 @@ import { DEFAULT_MODEL, MODEL_CONFIGS, readFiles, estimateRequestTokens, buildRe
 import { isKnownModel } from '../src/oracle/modelResolver.js';
 import { CHATGPT_URL } from '../src/browserMode.js';
 import { createRemoteBrowserExecutor } from '../src/remote/client.js';
+import { createGeminiWebExecutor } from '../src/gemini-web/index.js';
 import { applyHelpStyling } from '../src/cli/help.js';
 import { collectPaths, collectModelList, parseFloatOption, parseIntOption, parseSearchOption, usesDefaultStatusFilters, resolvePreviewMode, normalizeModelOption, normalizeBaseUrl, resolveApiModel, inferModelFromLabel, parseHeartbeatOption, parseTimeoutOption, mergePathLikeOptions, } from '../src/cli/options.js';
 import { copyToClipboard } from '../src/cli/clipboard.js';
@@ -26,6 +27,7 @@ import { shouldDetachSession } from '../src/cli/detach.js';
 import { applyHiddenAliases } from '../src/cli/hiddenAliases.js';
 import { buildBrowserConfig, resolveBrowserModelLabel } from '../src/cli/browserConfig.js';
 import { performSessionRun } from '../src/cli/sessionRunner.js';
+import { isMediaFile } from '../src/browser/prompt.js';
 import { attachSession, showStatus, formatCompletionSummary } from '../src/cli/sessionDisplay.js';
 import { formatCompactNumber } from '../src/cli/format.js';
 import { formatIntroLine } from '../src/cli/tagline.js';
@@ -85,7 +87,7 @@ program.hook('preAction', (thisCommand) => {
 });
 program
     .name('oracle')
-    .description('One-shot GPT-5.1 Pro / GPT-5.1 / GPT-5.1 Codex tool for hard questions that benefit from large file context and server-side search.')
+    .description('One-shot GPT-5.2 Pro / GPT-5.2 / GPT-5.1 Codex tool for hard questions that benefit from large file context and server-side search.')
     .version(VERSION)
     .argument('[prompt]', 'Prompt text (shorthand for --prompt).')
     .option('-p, --prompt <text>', 'User prompt to send to the model.')
@@ -110,18 +112,18 @@ program
     .addOption(new Option('--copy-markdown', 'Copy the assembled markdown bundle to the clipboard; pair with --render to print it too.').default(false))
     .addOption(new Option('--copy').hideHelp().default(false))
     .option('-s, --slug <words>', 'Custom session slug (3-5 words).')
-    .option('-m, --model <model>', 'Model to target (gpt-5.1-pro default; aliases to gpt-5.2-pro on API. Also gpt-5-pro, gpt-5.1, gpt-5.1-codex API-only, gpt-5.2, gpt-5.2-instant, gpt-5.2-pro, gemini-3-pro, claude-4.5-sonnet, claude-4.1-opus, or ChatGPT labels like "5.2 Thinking" for browser runs).', normalizeModelOption)
-    .addOption(new Option('--models <models>', 'Comma-separated API model list to query in parallel (e.g., "gpt-5.1-pro,gemini-3-pro").')
+    .option('-m, --model <model>', 'Model to target (gpt-5.2-pro default; also supports gpt-5.1-pro alias). Also gpt-5-pro, gpt-5.1, gpt-5.1-codex API-only, gpt-5.2, gpt-5.2-instant, gpt-5.2-pro, gemini-3-pro, claude-4.5-sonnet, claude-4.1-opus, or ChatGPT labels like "5.2 Thinking" for browser runs).', normalizeModelOption)
+    .addOption(new Option('--models <models>', 'Comma-separated API model list to query in parallel (e.g., "gpt-5.2-pro,gemini-3-pro").')
     .argParser(collectModelList)
     .default([]))
-    .addOption(new Option('-e, --engine <mode>', 'Execution engine (api | browser). Engine is preferred; --mode is a legacy alias. If omitted, oracle picks api when OPENAI_API_KEY is set, otherwise browser.').choices(['api', 'browser']))
+    .addOption(new Option('-e, --engine <mode>', 'Execution engine (api | browser). Browser engine: GPT models automate ChatGPT; Gemini models use a cookie-based client for gemini.google.com. If omitted, oracle picks api when OPENAI_API_KEY is set, otherwise browser.').choices(['api', 'browser']))
     .addOption(new Option('--mode <mode>', 'Alias for --engine (api | browser).').choices(['api', 'browser']).hideHelp())
     .option('--files-report', 'Show token usage per attached file (also prints automatically when files exceed the token budget).', false)
     .option('-v, --verbose', 'Enable verbose logging for all operations.', false)
     .addOption(new Option('--[no-]notify', 'Desktop notification when a session finishes (default on unless CI/SSH).')
     .default(undefined))
     .addOption(new Option('--[no-]notify-sound', 'Play a notification sound on completion (default off).').default(undefined))
-    .addOption(new Option('--timeout <seconds|auto>', 'Overall timeout before aborting the API call (auto = 60m for gpt-5.1-pro, 120s otherwise).')
+    .addOption(new Option('--timeout <seconds|auto>', 'Overall timeout before aborting the API call (auto = 60m for gpt-5.2-pro, 120s otherwise).')
     .argParser(parseTimeoutOption)
     .default('auto'))
     .addOption(new Option('--preview [mode]', '(alias) Preview the request without calling the model (summary | json | full). Deprecated: use --dry-run instead.')
@@ -182,6 +184,12 @@ program
     .addOption(new Option('--remote-token <token>', 'Access token for the remote `oracle serve` instance.'))
     .addOption(new Option('--browser-inline-files', 'Alias for --browser-attachments never (force pasting file contents inline).').default(false))
     .addOption(new Option('--browser-bundle-files', 'Bundle all attachments into a single archive before uploading.').default(false))
+    .addOption(new Option('--youtube <url>', 'YouTube video URL to analyze (Gemini web/cookie mode only; uses your signed-in Chrome cookies for gemini.google.com).'))
+    .addOption(new Option('--generate-image <file>', 'Generate image and save to file (Gemini web/cookie mode only; requires gemini.google.com Chrome cookies).'))
+    .addOption(new Option('--edit-image <file>', 'Edit existing image (use with --output, Gemini web/cookie mode only).'))
+    .addOption(new Option('--output <file>', 'Output file path for image operations (Gemini web/cookie mode only).'))
+    .addOption(new Option('--aspect <ratio>', 'Aspect ratio for image generation: 16:9, 1:1, 4:3, 3:4 (Gemini web/cookie mode only).'))
+    .addOption(new Option('--gemini-show-thoughts', 'Display Gemini thinking process (Gemini web/cookie mode only).').default(false))
     .option('--retain-hours <hours>', 'Prune stored sessions older than this many hours before running (set 0 to disable).', parseFloatOption)
     .option('--force', 'Force start a new session even if an identical prompt is already running.', false)
     .option('--debug-help', 'Show the advanced/debug option set and exit.', false)
@@ -512,18 +520,13 @@ async function runRootCommand(options) {
     const isCodex = primaryModelCandidate.startsWith('gpt-5.1-codex');
     const isClaude = primaryModelCandidate.startsWith('claude');
     const userForcedBrowser = options.browser || options.engine === 'browser';
-    const hasNonGptBrowserTarget = (engine === 'browser' || userForcedBrowser) &&
+    const isBrowserCompatible = (model) => model.startsWith('gpt-') || model.startsWith('gemini');
+    const hasNonBrowserCompatibleTarget = (engine === 'browser' || userForcedBrowser) &&
         (normalizedMultiModels.length > 0
-            ? normalizedMultiModels.some((model) => !model.startsWith('gpt-'))
-            : !resolvedModelCandidate.startsWith('gpt-'));
-    if (hasNonGptBrowserTarget) {
-        throw new Error('Browser engine only supports GPT-series ChatGPT models. Re-run with --engine api for Grok, Claude, Gemini, or other non-GPT models.');
-    }
-    if (isGemini && userForcedBrowser) {
-        throw new Error('Gemini is only supported via API. Use --engine api.');
-    }
-    if (isGemini && engine === 'browser') {
-        engine = 'api';
+            ? normalizedMultiModels.some((model) => !isBrowserCompatible(model))
+            : !isBrowserCompatible(resolvedModelCandidate));
+    if (hasNonBrowserCompatibleTarget) {
+        throw new Error('Browser engine only supports GPT and Gemini models. Re-run with --engine api for Grok, Claude, or other models.');
     }
     if (isClaude && engine === 'browser') {
         console.log(chalk.dim('Browser engine is not supported for Claude models; switching to API.'));
@@ -672,7 +675,11 @@ async function runRootCommand(options) {
         return;
     }
     if (options.file && options.file.length > 0) {
-        await readFiles(options.file, { cwd: process.cwd() });
+        const isBrowserMode = engine === 'browser' || userForcedBrowser;
+        const filesToValidate = isBrowserMode ? options.file.filter((f) => !isMediaFile(f)) : options.file;
+        if (filesToValidate.length > 0) {
+            await readFiles(filesToValidate, { cwd: process.cwd() });
+        }
     }
     const getSource = (key) => program.getOptionValueSource?.(key) ?? undefined;
     applyBrowserDefaultsFromConfig(options, userConfig, getSource);
@@ -698,6 +705,19 @@ async function runRootCommand(options) {
         };
         console.log(chalk.dim(`Routing browser automation to remote host ${remoteHost}`));
     }
+    else if (browserConfig && resolvedModel.startsWith('gemini')) {
+        browserDeps = {
+            executeBrowser: createGeminiWebExecutor({
+                youtube: options.youtube,
+                generateImage: options.generateImage,
+                editImage: options.editImage,
+                outputPath: options.output,
+                aspectRatio: options.aspect,
+                showThoughts: options.geminiShowThoughts,
+            }),
+        };
+        console.log(chalk.dim('Using Gemini web client for browser automation'));
+    }
     const remoteExecutionActive = Boolean(browserDeps);
     if (options.dryRun) {
         const baseRunOptions = buildRunOptions(resolvedOptions, {

package/dist/src/browser/actions/assistantResponse.js

@@ -218,6 +218,8 @@ async function isCompletionVisible(Runtime) {
         const ASSISTANT_SELECTOR = '${ASSISTANT_ROLE_SELECTOR}';
         const isAssistantTurn = (node) => {
           if (!(node instanceof HTMLElement)) return false;
+          const turnAttr = (node.getAttribute('data-turn') || node.dataset?.turn || '').toLowerCase();
+          if (turnAttr === 'assistant') return true;
           const role = (node.getAttribute('data-message-author-role') || node.dataset?.messageAuthorRole || '').toLowerCase();
           if (role === 'assistant') return true;
           const testId = (node.getAttribute('data-testid') || '').toLowerCase();
@@ -257,6 +259,12 @@ function normalizeAssistantSnapshot(snapshot) {
     if (!text.trim()) {
         return null;
     }
+    const normalized = text.toLowerCase();
+    // "Pro thinking" often renders a placeholder turn containing an "Answer now" gate.
+    // Treat it as incomplete so browser mode keeps waiting (and can click the gate).
+    if (normalized.includes('answer now') && (normalized.includes('pro thinking') || normalized.includes('chatgpt said'))) {
+        return null;
+    }
     return {
         text,
         html: snapshot?.html ?? undefined,
@@ -295,10 +303,13 @@ function buildResponseObserverExpression(timeoutMs) {
     const CONVERSATION_SELECTOR = ${conversationLiteral};
     const ASSISTANT_SELECTOR = ${assistantLiteral};
     const settleDelayMs = 800;
+    const ANSWER_NOW_LABEL = 'answer now';
 
     // Helper to detect assistant turns - matches buildAssistantExtractor logic
     const isAssistantTurn = (node) => {
       if (!(node instanceof HTMLElement)) return false;
+      const turnAttr = (node.getAttribute('data-turn') || node.dataset?.turn || '').toLowerCase();
+      if (turnAttr === 'assistant') return true;
       const role = (node.getAttribute('data-message-author-role') || node.dataset?.messageAuthorRole || '').toLowerCase();
       if (role === 'assistant') return true;
       const testId = (node.getAttribute('data-testid') || '').toLowerCase();
@@ -330,6 +341,11 @@ function buildResponseObserverExpression(timeoutMs) {
         });
         observer.observe(document.body, { childList: true, subtree: true, characterData: true });
         stopInterval = setInterval(() => {
+          // Pro thinking can gate the response behind an "Answer now" button. Keep clicking it while present.
+          const answerNow = Array.from(document.querySelectorAll('button,span')).find((el) => (el?.textContent || '').trim().toLowerCase() === ANSWER_NOW_LABEL);
+          if (answerNow) {
+            dispatchClickSequence(answerNow.closest('button') ?? answerNow);
+          }
           const stop = document.querySelector(STOP_SELECTOR);
           if (!stop) {
             return;
@@ -382,9 +398,10 @@ function buildResponseObserverExpression(timeoutMs) {
           lastLength = refreshed.text?.length ?? lastLength;
         }
         const stopVisible = Boolean(document.querySelector(STOP_SELECTOR));
+        const answerNowVisible = Boolean(Array.from(document.querySelectorAll('button,span')).find((el) => (el?.textContent || '').trim().toLowerCase() === ANSWER_NOW_LABEL));
         const finishedVisible = isLastAssistantTurnFinished();
 
-        if (!stopVisible || finishedVisible) {
+        if ((!stopVisible && !answerNowVisible) || finishedVisible) {
           break;
         }
       }
@@ -407,6 +424,10 @@ function buildAssistantExtractor(functionName) {
     const ASSISTANT_SELECTOR = ${assistantLiteral};
     const isAssistantTurn = (node) => {
       if (!(node instanceof HTMLElement)) return false;
+      const turnAttr = (node.getAttribute('data-turn') || node.dataset?.turn || '').toLowerCase();
+      if (turnAttr === 'assistant') {
+        return true;
+      }
       const role = (node.getAttribute('data-message-author-role') || node.dataset?.messageAuthorRole || '').toLowerCase();
       if (role === 'assistant') {
         return true;
@@ -443,11 +464,13 @@ function buildAssistantExtractor(functionName) {
       }
       const messageRoot = turn.querySelector(ASSISTANT_SELECTOR) ?? turn;
       expandCollapsibles(messageRoot);
-      const preferred =
-        messageRoot.querySelector('.markdown') ||
-        messageRoot.querySelector('[data-message-content]') ||
-        messageRoot;
-      const text = preferred?.innerText ?? '';
+      const preferred = messageRoot.querySelector('.markdown') || messageRoot.querySelector('[data-message-content]');
+      if (!preferred) {
+        continue;
+      }
+      const innerText = preferred?.innerText ?? '';
+      const textContent = preferred?.textContent ?? '';
+      const text = innerText.trim().length > 0 ? innerText : textContent;
       const html = preferred?.innerHTML ?? '';
       const messageId = messageRoot.getAttribute('data-message-id');
       const turnId = messageRoot.getAttribute('data-testid');
@@ -462,7 +485,7 @@ function buildCopyExpression(meta) {
     return `(() => {
     ${buildClickDispatcher()}
     const BUTTON_SELECTOR = '${COPY_BUTTON_SELECTOR}';
-    const TIMEOUT_MS = 5000;
+    const TIMEOUT_MS = 10000;
 
     const locateButton = () => {
       const hint = ${JSON.stringify(meta ?? {})};
@@ -526,53 +549,62 @@ function buildCopyExpression(meta) {
     };
 
     return new Promise((resolve) => {
-      const button = locateButton();
-      if (!button) {
-        resolve({ success: false, status: 'missing-button' });
-        return;
-      }
-      const interception = interceptClipboard();
-      let settled = false;
-      let pollId = null;
-      let timeoutId = null;
-      const finish = (payload) => {
-        if (settled) {
+      const deadline = Date.now() + TIMEOUT_MS;
+      const waitForButton = () => {
+        const button = locateButton();
+        if (button) {
+          const interception = interceptClipboard();
+          let settled = false;
+          let pollId = null;
+          let timeoutId = null;
+          const finish = (payload) => {
+            if (settled) {
+              return;
+            }
+            settled = true;
+            if (pollId) {
+              clearInterval(pollId);
+            }
+            if (timeoutId) {
+              clearTimeout(timeoutId);
+            }
+            button.removeEventListener('copy', handleCopy, true);
+            interception.restore?.();
+            resolve(payload);
+          };
+
+          const readIntercepted = () => {
+            const markdown = interception.state.text ?? '';
+            return { success: Boolean(markdown.trim()), markdown };
+          };
+
+          const handleCopy = () => {
+            finish(readIntercepted());
+          };
+
+          button.addEventListener('copy', handleCopy, true);
+          button.scrollIntoView({ block: 'center', behavior: 'instant' });
+          dispatchClickSequence(button);
+          pollId = setInterval(() => {
+            const payload = readIntercepted();
+            if (payload.success) {
+              finish(payload);
+            }
+          }, 100);
+          timeoutId = setTimeout(() => {
+            button.removeEventListener('copy', handleCopy, true);
+            finish({ success: false, status: 'timeout' });
+          }, TIMEOUT_MS);
           return;
         }
-        settled = true;
-        if (pollId) {
-          clearInterval(pollId);
-        }
-        if (timeoutId) {
-          clearTimeout(timeoutId);
+        if (Date.now() > deadline) {
+          resolve({ success: false, status: 'missing-button' });
+          return;
         }
-        button.removeEventListener('copy', handleCopy, true);
-        interception.restore?.();
-        resolve(payload);
-      };
-
-      const readIntercepted = () => {
-        const markdown = interception.state.text ?? '';
-        return { success: Boolean(markdown.trim()), markdown };
+        setTimeout(waitForButton, 120);
       };
 
-      const handleCopy = () => {
-        finish(readIntercepted());
-      };
-
-      button.addEventListener('copy', handleCopy, true);
-      button.scrollIntoView({ block: 'center', behavior: 'instant' });
-      dispatchClickSequence(button);
-      pollId = setInterval(() => {
-        const payload = readIntercepted();
-        if (payload.success) {
-          finish(payload);
-        }
-      }, 100);
-      timeoutId = setTimeout(() => {
-        button.removeEventListener('copy', handleCopy, true);
-        finish({ success: false, status: 'timeout' });
-      }, TIMEOUT_MS);
+      waitForButton();
     });
   })()`;
 }

package/dist/src/browser/actions/attachments.js

@@ -214,8 +214,33 @@ export async function waitForAttachmentCompletion(Runtime, timeoutMs, expectedNa
         const { result } = await Runtime.evaluate({ expression, returnByValue: true });
         const value = result?.value;
         if (value && !value.uploading) {
-            const attached = new Set((value.attachedNames ?? []).map((name) => name.toLowerCase()));
-            const missing = expectedNormalized.filter((name) => !attached.has(name));
+            const attachedNames = (value.attachedNames ?? [])
+                .map((name) => name.toLowerCase().replace(/\s+/g, ' ').trim())
+                .filter(Boolean);
+            const matchesExpected = (expected) => {
+                const baseName = expected.split('/').pop()?.split('\\').pop() ?? expected;
+                const normalizedExpected = baseName.toLowerCase().replace(/\s+/g, ' ').trim();
+                const expectedNoExt = normalizedExpected.replace(/\.[a-z0-9]{1,10}$/i, '');
+                return attachedNames.some((raw) => {
+                    if (raw.includes(normalizedExpected))
+                        return true;
+                    if (expectedNoExt.length >= 6 && raw.includes(expectedNoExt))
+                        return true;
+                    if (raw.includes('…') || raw.includes('...')) {
+                        const escaped = raw.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+                        const pattern = escaped.replace(/\\…|\\\.\\\.\\\./g, '.*');
+                        try {
+                            const re = new RegExp(pattern);
+                            return re.test(normalizedExpected) || (expectedNoExt.length >= 6 && re.test(expectedNoExt));
+                        }
+                        catch {
+                            return false;
+                        }
+                    }
+                    return false;
+                });
+            };
+            const missing = expectedNormalized.filter((expected) => !matchesExpected(expected));
             if (missing.length === 0) {
                 if (value.state === 'ready') {
                     return;
@@ -223,6 +248,11 @@ export async function waitForAttachmentCompletion(Runtime, timeoutMs, expectedNa
                 if (value.state === 'missing' && value.filesAttached) {
                     return;
                 }
+                // If files are attached but button isn't ready yet, give it more time but don't fail immediately
+                if (value.filesAttached) {
+                    await delay(500);
+                    continue;
+                }
             }
         }
         await delay(250);
@@ -249,7 +279,11 @@ export async function waitForAttachmentVisible(Runtime, expectedName, timeoutMs,
     };
 
     const turns = Array.from(document.querySelectorAll('article[data-testid^="conversation-turn"]'));
-    const userTurns = turns.filter((node) => node.querySelector('[data-message-author-role="user"]'));
+    const userTurns = turns.filter((node) => {
+      const turnAttr = (node.getAttribute('data-turn') || node.dataset?.turn || '').toLowerCase();
+      if (turnAttr === 'user') return true;
+      return Boolean(node.querySelector('[data-message-author-role="user"]'));
+    });
     const lastUser = userTurns[userTurns.length - 1];
     if (lastUser) {
       const turnMatch = Array.from(lastUser.querySelectorAll('*')).some(matchNode);

package/dist/src/browser/actions/modelSelection.js

@@ -63,12 +63,41 @@ function buildModelSelectionExpression(targetModel) {
       .map((token) => normalizeText(token))
       .filter(Boolean);
     const targetWords = normalizedTarget.split(' ').filter(Boolean);
+    const desiredVersion = normalizedTarget.includes('5 2')
+      ? '5-2'
+      : normalizedTarget.includes('5 1')
+        ? '5-1'
+        : normalizedTarget.includes('5 0')
+          ? '5-0'
+          : null;
+    const wantsPro = normalizedTarget.includes(' pro') || normalizedTarget.endsWith(' pro') || normalizedTokens.includes('pro');
+    const wantsInstant = normalizedTarget.includes('instant');
+    const wantsThinking = normalizedTarget.includes('thinking');
 
     const button = document.querySelector(BUTTON_SELECTOR);
     if (!button) {
       return { status: 'button-missing' };
     }
 
+    const getButtonLabel = () => (button.textContent ?? '').trim();
+    const buttonMatchesTarget = () => {
+      const normalizedLabel = normalizeText(getButtonLabel());
+      if (!normalizedLabel) return false;
+      if (desiredVersion) {
+        if (desiredVersion === '5-2' && !normalizedLabel.includes('5 2')) return false;
+        if (desiredVersion === '5-1' && !normalizedLabel.includes('5 1')) return false;
+        if (desiredVersion === '5-0' && !normalizedLabel.includes('5 0')) return false;
+      }
+      if (wantsPro && !normalizedLabel.includes(' pro')) return false;
+      if (wantsInstant && !normalizedLabel.includes('instant')) return false;
+      if (wantsThinking && !normalizedLabel.includes('thinking')) return false;
+      return true;
+    };
+
+    if (buttonMatchesTarget()) {
+      return { status: 'already-selected', label: getButtonLabel() };
+    }
+
     let lastPointerClick = 0;
     const pointerClick = () => {
       if (dispatchClickSequence(button)) {
@@ -106,8 +135,46 @@ function buildModelSelectionExpression(targetModel) {
       }
       let score = 0;
       const normalizedTestId = (testid ?? '').toLowerCase();
-      if (normalizedTestId && TEST_IDS.some((id) => normalizedTestId.includes(id))) {
-        score += 1000;
+      if (normalizedTestId) {
+        if (desiredVersion) {
+          // data-testid strings have been observed with both dotted and dashed versions (e.g. gpt-5.2-pro vs gpt-5-2-pro).
+          const has52 =
+            normalizedTestId.includes('5-2') ||
+            normalizedTestId.includes('5.2') ||
+            normalizedTestId.includes('gpt-5-2') ||
+            normalizedTestId.includes('gpt-5.2') ||
+            normalizedTestId.includes('gpt52');
+          const has51 =
+            normalizedTestId.includes('5-1') ||
+            normalizedTestId.includes('5.1') ||
+            normalizedTestId.includes('gpt-5-1') ||
+            normalizedTestId.includes('gpt-5.1') ||
+            normalizedTestId.includes('gpt51');
+          const has50 =
+            normalizedTestId.includes('5-0') ||
+            normalizedTestId.includes('5.0') ||
+            normalizedTestId.includes('gpt-5-0') ||
+            normalizedTestId.includes('gpt-5.0') ||
+            normalizedTestId.includes('gpt50');
+          const candidateVersion = has52 ? '5-2' : has51 ? '5-1' : has50 ? '5-0' : null;
+          // If a candidate advertises a different version, ignore it entirely.
+          if (candidateVersion && candidateVersion !== desiredVersion) {
+            return 0;
+          }
+          // When targeting an explicit version, avoid selecting submenu wrappers that can contain legacy models.
+          if (normalizedTestId.includes('submenu') && candidateVersion === null) {
+            return 0;
+          }
+        }
+        const matches = TEST_IDS.filter((id) => id && normalizedTestId.includes(id));
+        if (matches.length > 0) {
+          // Prefer the most specific match (longest token) instead of treating any hit as equal.
+          // This prevents generic tokens (e.g. "pro") from outweighing version-specific targets.
+          const best = matches.reduce((acc, token) => (token.length > acc.length ? token : acc), '');
+          score += 200 + Math.min(900, best.length * 25);
+          if (best.startsWith('model-switcher-')) score += 120;
+          if (best.includes('gpt-')) score += 60;
+        }
       }
       if (normalizedText && normalizedTarget) {
         if (normalizedText === normalizedTarget) {
@@ -134,6 +201,14 @@ function buildModelSelectionExpression(targetModel) {
         }
         score -= missing * 12;
       }
+      // If the caller didn't explicitly ask for Pro, prefer non-Pro options when both exist.
+      if (wantsPro) {
+        if (!normalizedText.includes(' pro')) {
+          score -= 80;
+        }
+      } else if (normalizedText.includes(' pro')) {
+        score -= 40;
+      }
       return Math.max(score, 0);
     };
 
@@ -153,7 +228,7 @@ function buildModelSelectionExpression(targetModel) {
           }
           const label = getOptionLabel(option);
           if (!bestMatch || score > bestMatch.score) {
-            bestMatch = { node: option, label, score };
+            bestMatch = { node: option, label, score, testid, normalizedText };
           }
         }
       }
@@ -182,11 +257,25 @@ function buildModelSelectionExpression(targetModel) {
         const match = findBestOption();
         if (match) {
           if (optionIsSelected(match.node)) {
-            resolve({ status: 'already-selected', label: match.label });
+            resolve({ status: 'already-selected', label: getButtonLabel() || match.label });
             return;
           }
           dispatchClickSequence(match.node);
-          resolve({ status: 'switched', label: match.label });
+          // Submenus (e.g. "Legacy models") need a second pass to pick the actual model option.
+          // Keep scanning once the submenu opens instead of treating the submenu click as a final switch.
+          const isSubmenu = (match.testid ?? '').toLowerCase().includes('submenu');
+          if (isSubmenu) {
+            setTimeout(attempt, REOPEN_INTERVAL_MS / 2);
+            return;
+          }
+          // Wait for the top bar label to reflect the requested model; otherwise keep scanning.
+          setTimeout(() => {
+            if (buttonMatchesTarget()) {
+              resolve({ status: 'switched', label: getButtonLabel() || match.label });
+              return;
+            }
+            attempt();
+          }, Math.max(120, INITIAL_WAIT_MS));
           return;
         }
         if (performance.now() - start > MAX_WAIT_MS) {

package/dist/src/browser/actions/promptComposer.js

@@ -283,20 +283,28 @@ async function verifyPromptCommitted(Runtime, prompt, timeoutMs, logger) {
     const primarySelectorLiteral = JSON.stringify(PROMPT_PRIMARY_SELECTOR);
     const fallbackSelectorLiteral = JSON.stringify(PROMPT_FALLBACK_SELECTOR);
     const script = `(() => {
-    const editor = document.querySelector(${primarySelectorLiteral});
-    const fallback = document.querySelector(${fallbackSelectorLiteral});
-    const normalize = (value) => value?.toLowerCase?.().replace(/\\s+/g, ' ').trim() ?? '';
-    const normalizedPrompt = normalize(${encodedPrompt});
-    const normalizedPromptPrefix = normalizedPrompt.slice(0, 120);
-    const CONVERSATION_SELECTOR = ${JSON.stringify(CONVERSATION_TURN_SELECTOR)};
-    const articles = Array.from(document.querySelectorAll(CONVERSATION_SELECTOR));
-    const normalizedTurns = articles.map((node) => normalize(node?.innerText));
-    const userMatched = normalizedTurns.some((text) => text.includes(normalizedPrompt));
-    const prefixMatched =
-      normalizedPromptPrefix.length > 30 &&
-      normalizedTurns.some((text) => text.includes(normalizedPromptPrefix));
-    const lastTurn = normalizedTurns[normalizedTurns.length - 1] ?? '';
-    return {
+	    const editor = document.querySelector(${primarySelectorLiteral});
+	    const fallback = document.querySelector(${fallbackSelectorLiteral});
+	    const normalize = (value) => {
+	      let text = value?.toLowerCase?.() ?? '';
+	      // Strip markdown *markers* but keep content (ChatGPT renders fence markers differently).
+	      text = text.replace(/\`\`\`[^\\n]*\\n([\\s\\S]*?)\`\`\`/g, ' $1 ');
+	      text = text.replace(/\`\`\`/g, ' ');
+	      text = text.replace(/\`([^\`]*)\`/g, '$1');
+	      return text.replace(/\\s+/g, ' ').trim();
+	    };
+	    const normalizedPrompt = normalize(${encodedPrompt});
+	    const normalizedPromptPrefix = normalizedPrompt.slice(0, 120);
+	    const CONVERSATION_SELECTOR = ${JSON.stringify(CONVERSATION_TURN_SELECTOR)};
+	    const articles = Array.from(document.querySelectorAll(CONVERSATION_SELECTOR));
+	    const normalizedTurns = articles.map((node) => normalize(node?.innerText));
+	    const userMatched =
+	      normalizedPrompt.length > 0 && normalizedTurns.some((text) => text.includes(normalizedPrompt));
+	    const prefixMatched =
+	      normalizedPromptPrefix.length > 30 &&
+	      normalizedTurns.some((text) => text.includes(normalizedPromptPrefix));
+	    const lastTurn = normalizedTurns[normalizedTurns.length - 1] ?? '';
+	    return {
       userMatched,
       prefixMatched,
       fallbackValue: fallback?.value ?? '',