Thinking Probe Chrome extension security report

Writing is thinking made visible. Every sentence you put on the page is a claim about the world. Some claims are clear and well-supported. Others are vague, circular, unsupported, or quietly wrong. Most writers struggle to see these gaps because they are too close to the work. They know what they meant to say, so they assume they said it. What you need in those moments is not another spell-checker. You need a Socratic interlocutor: someone who reads carefully and asks: * “What do you mean by that?” * “What evidence supports this?” * “You’re assuming X here — is that actually true?” * “Who would disagree, and why?” Thinking Probe automates that function. It reads your writing like a rigorous intellectual sparring partner and returns the questions that most need asking right now. The result: tighter, more honest, more well-examined prose — not because AI polished your words, but because it pushed you to think more clearly. ⸻ How It Works 1. You write Open any Google Doc and start drafting. Thinking Probe runs quietly in the background. 2. You pause After a short pause, and only once there is enough text for context, Thinking Probe analyzes your recent writing. 3. Questions appear A sidebar opens on the right side of your screen with short, direct probe questions. Each question targets a specific passage in your document where your reasoning may need clarification, evidence, or refinement. 4. Passages are highlighted The relevant passages are marked with a subtle amber underline so you know exactly what each question refers to. 5. You respond Type your answer in the sidebar. You can write a sentence or a paragraph — whatever helps you think through the issue. 6. A revision is suggested Thinking Probe uses your answer to generate a revised version of the highlighted passage that incorporates your clarified thinking. 7. You decide Click Copy & Paste to move the suggested revision into your document, or dismiss it and revise the passage yourself. The full cycle — pause, question, response, revision — is designed to fit naturally into a normal writing session. ⸻ Key Features Socratic Probe Questions Thinking Probe does not give generic feedback like “add more evidence” or “improve clarity.” It asks specific questions about the text you just wrote. Examples: * “You say this happens ‘often’ — what evidence shows it is common?” * “What does ‘better’ mean here: better for whom, and by what measure?” * “This assumes readers already accept your premise. What if they don’t?” * “You’ve made three claims in a row. Which one actually matters most?” The questions are short, direct, and designed to feel less like a classroom prompt and more like a smart reader pushing back. Inline Passage Highlighting Each probe question is tied to a specific passage in your Google Doc. Thinking Probe highlights the relevant passage with a light amber underline, so the connection between question and text is always clear. Only one set of highlights is active at a time. When a new set of probes appears, old highlights clear automatically. Multiple Probes Per Pass Each analysis pass can surface questions across multiple paragraphs at once, so you can work through a small set of probes at your own pace instead of waiting for feedback one paragraph at a time. Smart Triggering Logic Thinking Probe does not fire every time you type. It waits for a genuine pause, requires enough new writing since the last analysis, and only activates once your document has enough text to provide meaningful context. This keeps feedback from interrupting your drafting flow. No Re-Surfacing Old Questions Thinking Probe tracks which paragraphs have already been probed during your session. It will not keep asking about the same passage again and again. Once your current paragraphs have been cycled through, it resets and starts fresh. High-Quality Revision Suggestions After you answer a probe, Thinking Probe suggests a rewritten version of the highlighted passage. The revision is designed to reflect your clarified thinking while matching your existing tone and style. You get only the revised text — no preamble, no explanation, no meta-commentary. Non-Intrusive Sidebar The sidebar is anchored to the right side of Google Docs. It does not cover your writing, block the toolbar, or interfere with the comments pane. You can collapse it when you want a cleaner writing space. No Account or API Key Required Thinking Probe works out of the box. You do not need to create an Anthropic account, manage an API key, or configure anything before writing. AI calls are handled in the background. Monthly Usage Allocation Each installation includes a monthly allocation of probe and edit calls. Usage is tracked locally and resets at the start of each calendar month. The allocation is designed to cover normal active writing sessions. ⸻ Who It’s For Thinking Probe is useful for anyone writing serious prose where the quality of the thinking matters. Students writing essays and papers Thinking Probe does not do the work for you. It does the opposite: it pushes you to clarify what you are claiming, whether your evidence supports it, and where your reasoning needs work. Researchers and academics Academic writing often contains assumptions and logic that feel obvious to the writer because they have been living with the material for months. Thinking Probe offers an outside reader’s pressure at the draft stage. Knowledge workers writing under time pressure Fast first drafts often skip over the weakest links in an argument. Thinking Probe helps surface those links before you send the document to your manager, team, or client. Journalists and essayists Opinion writing and long-form analysis depend on the strength of the reasoning underneath the prose. Thinking Probe helps stress-test your claims before readers do. Founders and operators Strategy docs, memos, investor updates, and internal narratives all get read by people who will push back. Thinking Probe helps you find the soft spots first. Anyone who wants to write more rigorously If you have ever finished a draft and felt that it could be sharper, more honest, or more precise, Thinking Probe turns that unease into useful questions. ⸻ What Thinking Probe Is Not Thinking Probe is not a grammar checker. It does not focus on spelling, punctuation, passive voice, sentence length, or readability scores. Thinking Probe is not an AI writer. It does not autocomplete your draft, summarize your document, or generate content from scratch. Thinking Probe is not a style enforcer. It will not tell you that your sentences are too long, your vocabulary is too advanced, or your structure is non-standard. Thinking Probe is not a note-taking or outlining tool. It does not store your writing history or maintain memory across sessions. Thinking Probe does one thing: it reads what you wrote and asks the questions most likely to make the thinking stronger. ⸻ Privacy and Data Thinking Probe sends excerpts of your Google Doc text to Claude, Anthropic’s AI model, through a secure backend service for analysis. The extension does not store your document text, log your writing, or share your content with any third party beyond what is necessary to generate probe questions and revision suggestions. The extension stores only: 1. Your monthly usage count 2. A record of which paragraphs have already been probed during the current session No writing content is stored locally or remotely between sessions. You can remove locally stored data at any time by uninstalling the extension. ⸻ Setup 1. Install Thinking Probe. 2. Open any Google Doc. 3. If prompted, enable screen reader support in Google Docs: Tools → Accessibility settings → Turn on screen reader support 4. Start writing. 5. The sidebar appears automatically when Thinking Probe has enough context to generate probes. No accounts. No API keys. No configuration required. ⸻ Frequently Asked Questions Do I need an Anthropic account or API key? No. Thinking Probe works out of the box. You do not need an Anthropic account, API key, or subscription. Is it free? Yes. Each installation includes a monthly allocation of probe and edit calls at no charge. The allocation resets at the start of each calendar month. Why do I need to enable screen reader support in Google Docs? Google Docs renders document text in a way that is not normally accessible to browser extensions. Enabling screen reader support makes the text available through Google Docs’ accessibility layer, which allows Thinking Probe to read the document. This is a one-time setup step. It does not change the appearance of your document. Does it work on all Google Docs? Thinking Probe works on Google Docs in Chrome when screen reader support is enabled. It does not currently support Google Sheets, Google Slides, or other Google Workspace apps. Does it work in Notion, Gmail, or other writing tools? Not in version 1. Google Docs is the only supported editor for now. How is this different from asking ChatGPT or Claude to review my writing? The difference is timing, focus, and integration. Using a chatbot requires you to leave your document, paste in your writing, decide what kind of feedback you want, and parse a long response. Thinking Probe stays inside Google Docs and surfaces targeted questions automatically while you write. It does not review everything. It asks about the specific passages where your reasoning needs pressure. Can I hide the sidebar? Yes. The sidebar has a collapse toggle. When collapsed, Thinking Probe can continue running in the background and queue new probes until you expand it again. What if the questions feel irrelevant? Dismiss the question and keep writing. You can also refresh to request a new set of probes. The quality of questions usually improves as your draft becomes more specific and argumentative. Does the extension read my entire document? No. Thinking Probe analyzes only recent paragraphs around your current writing context. It does not send your full document, document title, revision history, or Google account data. What if Google Docs changes and the extension breaks? Google Docs can change its internal editor structure without notice. If that happens, Thinking Probe may temporarily lose the ability to extract text or apply highlights. An update will be released to restore compatibility.