Privacy Policy

Citation Gap Finder (Chrome extension)

What we access: Citation Gap Finder reads the text content of your open document in Google Docs or Overleaf only when you click "Run Check." We do not read content automatically or in the background.

How we read your document:

• Overleaf: we read text directly from the editor in your browser. Your document text never leaves your browser.
• Google Docs: because the modern Google Docs editor renders text to a canvas (no readable DOM), we ask Google's own plain-text export endpoint (https://docs.google.com/document/d/.../export?format=txt) for your document text. This request is sent to Google — who already hosts your document — using your existing browser session. Your document text never reaches our servers and is never seen by any third party. This is the same operation Google offers under File → Download → Plain Text.

What we send to our servers:

• License key validation: when you have a license, we send only your license key to our validation server (a Cloudflare Worker). We do not send any document content.

Citation Gap Finder makes no other network calls. DOI checking is performed entirely in your browser using a strict regex-validation step against the DOIs in your reference list — we do not call CrossRef or any third-party DOI registry from the extension.

What we store:

• Locally in your browser (IndexedDB): your license key, your trial status, and a count of issues found in your last 50 checks. No document content is stored.
• On our servers: your email address (only if you purchase a license), your license key, and your subscription status.

Payment processing: Citation Gap Finder subscriptions are processed by Stripe. When you purchase, you provide your payment information directly to Stripe — we never see or store your card details. See Stripe's Privacy Policy.

Activation emails: we send your license key to the email address you provide at checkout via Resend (our transactional email provider). The sending domain is support@gradsummit.com. See Resend's Privacy Policy.

Research Quote Capture (Chrome extension)

What we access. When you press Ctrl+Shift+Q (Cmd+Shift+Q on Mac), the extension reads the text you have selected on the current tab, the page's URL, its <title>, and a small set of citation-related <meta> tags (such as citation_doi, citation_title, citation_author). Reading happens only at the moment you press the shortcut, only on the tab you have focused, and only via Chrome's activeTab permission. The extension does not run on any page until you invoke the shortcut.

How we read your page.

• We call window.getSelection() to read the highlighted text.
• We read document.location.href and document.title.
• We read citation meta tags (citation_doi, citation_title, citation_author, citation_publication_date, citation_journal_title, dc.identifier, dc.creator) and JSON-LD ScholarlyArticle data, when present.
• We do not read form fields, cookies, localStorage, sessionStorage, or any other tab data.

What we send to our servers.

• Your license key (string only) — sent to research-quote-capture-api.muhraf-livify.workers.dev for activation, validation, and revocation checks.
• For Pro users only: the detected DOI string — sent to api.openalex.org to fetch citation metadata. OpenAlex is a public, free academic-citation API operated by OurResearch. The DOI string contains no personally identifying information.
• We never send the selected text, the page URL, the page title, or any other page content to any server.

What we store.

• Locally on your device only: all captured quotes, including the selected text, page URL, page title, timestamp, DOI, and any citation metadata returned by OpenAlex. Stored in the browser's IndexedDB. Deleted when you uninstall the extension or clear browser data for the extension.
• On our license server: license key, the email you used at checkout (so we can resend the key on request), purchase timestamp, subscription status, and a 7-day cache of validation results. No quote data, no URLs, no page content.

Payment processing. Subscriptions are processed by Stripe. We never see your card number. Stripe's privacy policy.

Activation emails. License keys are delivered by Resend. Resend's privacy policy.

Methods Section Sanity Checker (Chrome extension)

What we access: Methods Section Sanity Checker reads the methods section of the Google Doc you have open only when you click Scan methods section. On the Paste tab, only the text you type or paste into the textarea inside the side panel.

How we read your document:

• Google Docs: via the standard ?format=txt export endpoint (https://docs.google.com/document/d/.../export?format=txt) — the same operation Google offers under File → Download → Plain Text. The request is sent to Google — who already hosts your document — using your existing browser session. Your document text never reaches our servers and is never seen by any third party.
• Paste tab (Overleaf, Word, Pages, Scrivener, anywhere else): only the text you paste into the textarea inside the extension's side panel. Stays in your browser process; nothing is scraped, nothing is sent to our servers. v1 ships Google Docs as the only one-click integration; first-class Overleaf integration arrives in v1.1.

What we send to our servers:

• License key validation: when you have a license, we send only your license key to our validation server (a Cloudflare Worker at methods-checker-api.muhraf-livify.workers.dev) — once at activation and once every seven days thereafter. We do not send any document content.

Methods Section Sanity Checker makes no other network calls. The detection engine runs as pure regex over curated keyword libraries entirely inside your browser — we do not call any AI, LLM, or external API to score your text.

What we store:

• Locally on this device only (in chrome.storage.local and IndexedDB): your license key, your chosen default guideline, your "Mark as addressed" overrides per document URL, and a count of items found in your last 50 scans (counts only — never document text). Deleted when you uninstall the extension or clear browser data for the extension.
• In Chrome's sync storage (in chrome.storage.sync): your trial start timestamp, and only that. Chrome syncs this to your Google account so the 7-day trial timer follows you across the devices you use Chrome on; the value is encrypted by Chrome in transit and is not visible to us. It never reaches our servers. If you have Chrome sync turned off, the timestamp stays on this device only.
• On our servers (Cloudflare Workers KV): your license key, the email you used at checkout, your subscription status, and timestamps for activation and renewal. No document content. No URLs. No usage patterns.

Payment processing. Methods Section Sanity Checker subscriptions are processed by Stripe. When you purchase, you provide your payment information directly to Stripe — we never see or store your card details. See Stripe's Privacy Policy.

Activation emails. License keys are delivered by Resend. The sending domain is support@gradsummit.com. See Resend's Privacy Policy.

Research Assistant (Chrome extension)

What we access: Research Assistant only ever reads the text you type, paste, or upload (as a PDF) into its side panel, and only when you run a feature. It does not read the pages you browse, and it has no access to your documents in any editor.

How it processes your draft: entirely inside your browser. The journal matcher runs a small AI sentence-embedding model locally (via WebAssembly/WebGPU); PDFs are converted to text locally (pdf.js); the drafts-library overlap check (MinHash) runs locally. Your manuscript text is never sent to GradSummit or any third party. There is no backend that stores drafts.

What leaves your device, and where it goes:

• One-time model download: the embedding model files are downloaded once from the Hugging Face CDN (huggingface.co and its asset hosts) and then cached locally. These are model data files, not your content.
• "What should I cite?" search: short, anonymous phrases sampled from your draft (not your full draft) are sent to the public bibliographic APIs eutils.ncbi.nlm.nih.gov (PubMed), api.openalex.org (OpenAlex), and api.crossref.org (Crossref) to find prior work. These are public APIs operated by NCBI, OurResearch, and Crossref respectively — the same queries a researcher would type into a search box. If you add your own free NCBI API key, it is sent only to NCBI.
• Optional product-updates email: only if you tick "product updates" and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No manuscript content is ever included. You never need to provide an email to use any feature.

What we store:

• Locally in your browser only (IndexedDB and chrome.storage.local): your drafts library (full text and its MinHash signatures), a summary of your last 50 results, the cached journal index, and your settings (an optional NCBI API key and an optional email). No manuscript content is stored on our servers. Deleted when you uninstall the extension, clear its data, or use the in-app "Clear data" controls.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Research Assistant is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Research Assistant is a discovery and self-check assistant — it is not a plagiarism checker and contains no AI-content detector.

Thesis Progress Reality Check (Chrome extension)

What we access: Thesis Progress Reality Check only ever reads what you type into its own side panel and welcome tab — your program details, thesis structure, weekly word counts, status changes, obstacle categories, commitments, and strategic notes. It does not read the pages you browse, has no access to your documents (Word, Google Docs, Overleaf, or any editor), runs no content scripts, and does not auto-detect word counts.

How it processes your entries: entirely inside your browser. The three pace projections, the pattern-detection observations, and the timeline visualisation are all computed deterministically on your device from values you enter yourself. There is no AI runtime; no LLM, no inference, no model download. Your thesis data is never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you enter an email in Settings (or on the welcome page) and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No thesis content, word counts, statuses, weekly reviews, obstacles, commitments, or notes are ever included. You never need to provide an email to use any feature, and this is the only network request the extension makes.

What we store:

• Locally in your browser only (IndexedDB and chrome.storage.local): all program details, thesis structure, weekly snapshots, obstacle categories, commitments, strategic decisions, and your settings (including the optional weekly-reminder schedule and an optional opt-in email). No thesis content is stored on our servers. Deleted when you uninstall the extension, clear its data, or use the in-app "Delete all data" control in Settings.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Notifications and reminders: the optional weekly review reminder uses Chrome's built-in alarm and notification system. The alarm fires locally; we do not transmit reminder content to any server.

Wellbeing note: this extension is not a mental-health tool. It does not collect, infer, or transmit any mental-health information. The "obstacles" categories you can select during a weekly review (including a "mental health" option) are stored only on your device.

Payment processing: none. Thesis Progress Reality Check is free, with no subscription, no licence key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Thesis Progress Reality Check is a structured tracker — it is not a predictor of success or failure, and not a mental-health tool.

Research Decision Journal (Chrome extension)

What we access: Research Decision Journal only ever reads what you type into its own workspace — your decisions, alternatives, reasoning, assumptions, reviews, and notes. It does not read the pages you browse, has no access to your documents in any editor, and runs no content scripts.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing evaluates your reasoning, suggests alternatives, codes your outcomes, or generates prompts. Review schedules and calibration metrics are computed deterministically on your device from values you enter and outcomes you code yourself. Your decision content is never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you tick "product updates" and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No decision content is ever included. You never need to provide an email to use any feature, and this is the only network request the extension makes.

What we store:

• Locally in your browser only (chrome.storage.local): all of your decisions, reviews, settings, and an optional email. No decision content is stored on our servers. Deleted when you uninstall the extension or use the in-app "Clear all data" control. Your private "internal note" fields are excluded from every export.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Research Decision Journal is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Research Decision Journal is a workspace for structured decision recording — it is not an AI coach and does not send you reminders by email, push notification, or calendar.

Supervisor Meeting Brief Generator (Chrome extension)

What we access: Supervisor Meeting Brief Generator only ever reads what you type into its own workspace — your meeting dates, supervisor profiles, pre-meeting briefs (status, wins, reading, questions, decisions, blockers, milestones), post-meeting captures (decisions made, questions answered, action items, references suggested, recognition received, feedback, disagreements, themes), and your private notes. It does not read the pages you browse, has no access to your calendar or documents in any editor, and runs no content scripts. It declares no <all_urls> permission and no host permissions other than the sign-up endpoint described below.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing generates your status bullets, suggests decisions, summarises captures, or infers themes. The speaking-time meter, action-item staleness detection, theme deduplication, and full-text search are all deterministic browser code. Your meeting content is never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you tick "Subscribe to product updates" and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No brief content, capture content, supervisor names, action items, themes, or notes are ever included. You never need to provide an email to use any feature, and this is the only network request the extension can make.

What we store:

• Locally in your browser only (chrome.storage.local and IndexedDB): all of your meetings, supervisor profiles, briefs, captures, action items, themes, settings, and an optional email. No meeting content is stored on our servers. Deleted when you uninstall the extension or use the in-app "Erase all data" control. The "Private notes" fields on every brief and capture are excluded from PDF, email, and Markdown exports.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Supervisor Meeting Brief Generator is free, with no subscription, no license key, and no account.

Notifications: the optional 1–168 hour pre-meeting reminder uses Chrome's built-in notification system. The notification content is generated locally based on your own scheduled meetings; nothing is sent to a server when a reminder fires.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Supervisor Meeting Brief Generator is a workspace for structured supervisor-meeting preparation and capture — it is not an AI coach, not a generic meeting-notes app, and does not read or interact with any web page you have open.

Lit Review Synthesizer (Chrome extension)

What we access: Lit Review Synthesizer reads the open paper page only when you click its toolbar icon on one of four supported sites (pubmed.ncbi.nlm.nih.gov, www.biorxiv.org / biorxiv.org, arxiv.org, scholar.google.com) to pull the paper's title, authors, year, journal, and DOI. On Google Scholar, a small "Capture in LRS" link is injected next to each result so you can stash a paper's metadata for the popup; clicking it does not transmit anything. The extension has no <all_urls> permission and runs no content scripts on any other website.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing reads your cards, generates synthesis paragraphs, or codes your stances. The matrix view, Synthesis Insights flags, and Chapter Scaffold export are computed by deterministic string templates and conditional logic over the cards you entered. Your synthesis card content is never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional OpenAlex enrichment: when you trigger enrichment on a paper (per-paper, or via the bulk-enrich action in Settings), the extension sends only that paper's DOI to OpenAlex's public API (api.openalex.org) and receives back the abstract, citation count, and concept tags, which are cached locally. OpenAlex is a public bibliographic API operated by OurResearch. Your synthesis card content is never included.
• Optional product-updates email: only if you tick "Email me GradSummit product updates" and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No synthesis card content, paper data, notes, or research question is ever included. You never need to provide an email to use any feature.

What we store:

• Locally in your browser only (chrome.storage.local): your research question, discipline, custom methodology list, all captured papers and their synthesis cards (research question, finding, methodology, stance, optional rigor / scope / relevance, notes), cached OpenAlex enrichment data, and an optional email. No synthesis or paper content is stored on our servers. Deleted when you uninstall the extension or use the in-app "Wipe everything" control in Settings.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Lit Review Synthesizer is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Lit Review Synthesizer is a workspace for the structural synthesis of a literature review — it is not an AI synthesizer and does not write your paragraphs, suggest stances, or replace your supervisor's judgment.

Hypothesis Stress Test (Chrome extension)

What we access: Hypothesis Stress Test only ever reads what you type into its own workspace — your hypothesis statements, written defences, evidence citations, residual vulnerabilities, internal notes, follow-up responses, and version history. It does not read the pages you browse, has no access to your documents in any editor, and runs no content scripts.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing evaluates your defences, suggests counter-arguments, grades your reasoning, or generates attacks. The catalog of 18 attacks is curated authorial work shipped in the extension bundle; the adversarial follow-ups are deterministic predicates over what you wrote; engagement metrics are deterministic counts. Your hypothesis content and defence text are never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you tick "I'd like occasional product updates" and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No hypothesis content, defence text, evidence, or notes are ever included. You never need to provide an email to use any feature, and this is the only network request the extension makes.

What we store:

• Locally in your browser only (chrome.storage.local): all of your hypotheses, defences, follow-up responses, settings, and an optional email. No hypothesis or defence content is stored on our servers. Deleted when you uninstall the extension or use the in-app "Clear all data" control. Your private "internal note" fields are excluded from every export.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Hypothesis Stress Test is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Hypothesis Stress Test is a workbook for adversarial self-examination — it is not an AI critic and does not evaluate the quality of your defences. The catalog is fixed authorial content; you do the substantive thinking.

Qual Prep Tracker (Chrome extension)

What we access: Qual Prep Tracker reads the open paper page only when you click its toolbar icon on one of four supported sites (pubmed.ncbi.nlm.nih.gov, www.biorxiv.org / biorxiv.org, arxiv.org, scholar.google.com) to pull the paper's title, authors, year, DOI, and journal so the capture form pre-fills. The extension has no <all_urls> permission, runs no content scripts on any other website, and never reads any other page.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing summarizes your papers, recommends what to read, scores your coverage, or judges whether you're exam-ready. The coverage map, gap report tiers, and reading-priority checklist are computed by deterministic logic over the papers and depth tags you entered yourself. Your paper content, notes, framework customizations, and settings are never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional OpenAlex enrichment: when you trigger enrichment on a paper (per-paper from the papers list, or via the auto-enrich setting), the extension sends only that paper's DOI to OpenAlex's public API (api.openalex.org) and receives back the abstract, citation count, and concept tags, which are cached locally. OpenAlex is a free, open scholarly database operated by OurResearch (a non-profit). The paper title, authors, journal, your notes, your area/sub-topic tags, your depth marking, your framework, and your exam date are never included.
• Optional product-updates email: only if you tick "Email me GradSummit product updates" and click Subscribe, the extension sends only the email address you typed (plus a tag naming the extension) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No paper content, notes, framework, exam date, or other settings are ever included. You never need to provide an email to use any feature.

What we store:

• Locally in your browser only (chrome.storage.local): your chosen field framework and any customizations (added or hidden areas / sub-topics), every captured paper (title, authors, year, DOI, journal, URL, source platform, area + sub-topic tag, depth, notes, capture timestamp), cached OpenAlex enrichment data for papers you enriched, your exam date, weekly-hours target, sub-specialty, any field requests you saved, and an optional email. None of this is stored on our servers. Deleted when you uninstall the extension or use the in-app "Wipe local data" control in Settings → Data.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Qual Prep Tracker is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Qual Prep Tracker is a workspace for tracking literature breadth before a qualifying exam — it is not an AI tutor, does not predict exam outcomes, and does not recommend specific papers. The framework is structure; the reading is yours.

Defense Prep Workbench (Chrome extension)

What we access: Defense Prep Workbench only ever reads what you type into its own workspace — your thesis title, discipline, defense date, vulnerability articulations, examiner research notes, draft answers, answer notes, practice history, internal notes, and any free-text fields. It does not read the pages you browse, has no access to your documents in any editor, and runs no content scripts. It declares only the storage permission and a single host_permission for the GradSummit sign-up Worker described below.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing generates your answers, evaluates your reasoning, researches your examiners, or predicts the specific questions you will be asked. The curated library of ~150 anticipated questions across 10 categories is static Markdown content shipped in the extension bundle. The schedule calculator, spaced-review hints, completeness indicators, and exports are all computed by deterministic string templates and rules over the entries you supply. Your defense prep content is never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you tick "Email me occasional updates" and save, the extension sends only the email address you typed (plus a defense-prep-workbench tag) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No defense prep content, vulnerability text, examiner notes, draft answers, practice history, or internal notes are ever included. You never need to provide an email to use any feature, and this is the only network request the extension makes.

What we store:

• Locally in your browser only (chrome.storage.local): your defense metadata, schedule state, vulnerabilities, examiner profiles, question bank with draft answers and notes, practice history, settings, and an optional email. No defense prep content is stored on our servers. Deleted when you uninstall the extension or use the in-app "Reset all data" control in Settings. Your private "internal note" fields are excluded from every export.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Defense Prep Workbench is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Defense Prep Workbench is a structured workbench for defense preparation — it is not an AI coach, does not generate your answers or evaluate your reasoning, and does not guarantee a defense outcome. The structure, curated content, and captured artifacts are what the tool provides; the substantive work is yours.

Defense Preparation Coach (Chrome extension)

What we access: Defense Preparation Coach only ever reads what you type into its own side-panel workspace — your defense date, thesis title, discipline, the questions you add across the eight categories, your draft answers, vulnerability notes (with severity and mitigation), examiner profiles (with expertise and likely angles), and the practice ratings you record after each rehearsal. It does not read the pages you browse, has no access to your documents in any editor, and runs no content scripts. It declares only the sidePanel and storage permissions and a single host_permission for the GradSummit sign-up Worker described below.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing generates your questions, drafts your answers, or evaluates your reasoning. The speaking-time estimator is a plain word count divided by a 130-words-per-minute constant; the weighted practice queue is a deterministic ranking over recency and your own self-ratings. There is no microphone access, no audio recording, and no transcription, ever. Your defense prep content is never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you tick "Get occasional defense-prep tips" and submit, the extension sends only the email address you typed (plus a defense-preparation-coach tag) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No question text, draft answer text, vulnerability note, examiner detail, or practice rating is ever included. You never need to provide an email to use any feature, and this is the only network request the extension makes.

What we store:

• Locally in your browser only (IndexedDB and chrome.storage.local): your defense date, thesis title, discipline, question bank, draft answers, vulnerability notes, examiner profiles, practice session log, and an optional email. No defense prep content is stored on our servers. Deleted when you uninstall the extension or clear the extension's storage in your browser.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Payment processing: none. Defense Preparation Coach is free, with no subscription, no license key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Defense Preparation Coach is a rehearsal coach for time-bounded defense practice — it is not an AI examiner, does not generate your questions or answers, does not record or analyze audio, and does not guarantee a defense outcome. The structure, the timing, and the self-rating loop are what the tool provides; the substantive work is yours.

Feedback Triage (Chrome extension)

What we access: Feedback Triage only ever reads what you type or paste into its own workspace — the source feedback you paste (supervisor emails, peer reviewer reports, committee notes, editor decision letters, grant reviewer comments, reading-group notes), the items you extract from it (verbatim text, type, severity, underlying concern, specific action, success criteria, effort estimate, status, decided-not-to-address rationale, and a private internal note), and your session metadata (titles, project tags, optional source-type fields like journal name, deadline, decision). It does not read the pages you browse, has no access to your documents in any editor, runs no content scripts, and declares no <all_urls> or per-site host permissions. It declares only the storage permission and a single host_permission for the GradSummit sign-up Worker described below.

How it processes your entries: entirely inside your browser. There is no AI runtime — nothing classifies the type, suggests the severity, infers the underlying concern, drafts your response paragraphs, or summarises anything. The highlight-and-extract flow, severity colouring, output templates and Markdown exports are deterministic browser code. Your pasted feedback, items, concerns, actions, and internal notes are never sent to GradSummit or any third party.

What leaves your device, and where it goes:

• Optional product-updates email: only if you tick "I'd like occasional product updates" and click Subscribe, the extension sends only the email address you typed (plus a feedback-triage tag) to our sign-up endpoint — a Cloudflare Worker at gradsummit-subscribe.muhraf-livify.workers.dev — which records it and adds it to our mailing list via Resend. No pasted feedback, verbatim quote, classification, underlying-concern interpretation, action plan, internal note, or session metadata is ever included. You never need to provide an email to use any feature, and this is the only network request the extension makes.

What we store:

• Locally in your browser only (chrome.storage.local): all of your triage sessions, items, classifications, action plans, source-type metadata, type-list customisations, and an optional email. No feedback content is stored on our servers. Deleted when you uninstall the extension or use the in-app "Delete all data" control. Your private "internal note" fields on items are excluded from every export — verified by an automated test in the build's compatibility battery.
• On our servers (only if you opt in to updates): your email address, the extension you signed up from, and a timestamp — stored in Cloudflare D1 and Resend so we can email you occasional updates.

Reviewer confidentiality: many journals' reviewer agreements prohibit redistribution of reviewer comments. Pasting reviewer comments into Feedback Triage is not a transfer to a third party — the content remains in your browser, on your device, and is not transmitted to GradSummit, the journal, or anyone else. The legal interpretation of any specific journal's policy is between you and the journal.

Payment processing: none. Feedback Triage is free, with no subscription, no licence key, and no account.

Email: product-update emails (and your sign-up) are handled by Resend, our email provider. See Resend's Privacy Policy. You can unsubscribe at any time via the link in any email or by contacting support@gradsummit.com.

Feedback Triage is a thinking tool for processing received critique — it is not an AI that decodes reviewer comments for you, does not auto-classify or auto-prioritise, and does not guarantee a revision outcome. The structure, the prompts, and the exports are what the tool provides; the judgment is yours.

Books (Mastering Research series)

Books in the Mastering Research series are sold and delivered by Amazon. Your purchase, billing, and delivery information is handled entirely by Amazon under their privacy policy. We do not collect or store any data about book purchases on this website.

This website (gradsummit.com)

The website itself uses Google Analytics (GA4) to count visits and understand which pages are useful. The data is aggregated and we do not associate page views with individual identities. You can opt out of Google Analytics by installing the official browser opt-out.

We do not use marketing cookies, ad-tracking pixels, or third-party trackers beyond Google Analytics.