Lit Review Synthesizer

Lit Review Synthesizer scaffolds the structural synthesis work of a literature review — the part supervisors mark students down for never doing. You write one structured card per paper (research question, finding, methodology, your stance). The cards roll up into a matrix you can filter, sort, and group. When it's time to write your chapter, you export a Markdown scaffold with the structural work done — sections grouped by methodology and stance, with gaps surfaced automatically, and bracketed prompts where your synthesis paragraphs go. Your synthesis cards stay in your browser. No AI runtime, no telemetry, no account — and the synthesis paragraphs (the part you're being graded on) remain yours. ★ WHAT IT DOES 1. CAPTURE — one-click metadata from PubMed, bioRxiv, arXiv, and Google Scholar (with a "Capture in LRS" link injected next to each result). Manual entry for everything else — first-class peer, not a fallback. 2. CARD — per paper: paper's research question, core finding, methodology (customisable pick-list with discipline defaults), your stance toward your hypothesis (Supports / Contradicts / Nuances / Tangential / Undetermined). Optional rigor, scope, relevance, notes. 3. MATRIX — full-page view: search, filter, sort, group. Stance as a colored pill; rigor as 3-dot indicator. Click any row to expand; edit any field in a modal. 4. SYNTHESIS INSIGHTS — deterministic flags (no AI) that surface confirmation bias, methodology dominance, weak rigor concentration, narrow time windows, missing nuances, dated sources. 5. CHAPTER SCAFFOLD EXPORT — a full Markdown outline with Introduction (auto methodology-mix sentence), Methodological Approaches grouped by approach, Findings — Where the Field Agrees, Findings — Where the Field Is Contested, Identified Gaps populated from active Insights, and full BibTeX bibliography. Bracketed [Synthesize: …] prompts mark where YOUR paragraphs go. 6. EXPORTS — BibTeX (Zotero, LaTeX), CSV, JSON (backup + import), Markdown matrix. All respect the current filters. 7. OPENALEX ENRICHMENT — opt-in. Sends only the DOI; returns abstract, citation count, concepts. Cached locally. ★ PRIVATE BY DESIGN - All synthesis cards, the research question, notes, and settings live in chrome.storage.local. Nothing is uploaded. - No AI runtime, no telemetry, no analytics, no account. - Two scoped outbound calls and both are opt-in: OpenAlex enrichment (DOI only, triggered by you) and product-updates sign-up (only your email, only if you tick the box). - No <all_urls>. Host permissions are limited to the four capture sites, the OpenAlex API, and the GradSummit sign-up Worker — that's it. ★ FREE - Every feature, for everyone — no subscription, no tiers, no license key, no account. ★ WHAT THIS IS NOT - Not an AI synthesizer. It does not write your paragraphs, suggest stances, or auto-categorize. That separation is the whole point. - Not a systematic review tool. No PRISMA, no risk-of-bias instruments, no dual screening. Use Covidence / Rayyan / DistillerSR for that workflow. - Not a citation manager. Use Zotero; the BibTeX export is the bridge. - Not a PDF reader or annotator. Out of scope for v1. ★ HOW IT WORKS UNDER THE HOOD - Plain JavaScript ES modules, no build step. - Manifest V3 service worker; scoped host_permissions (no <all_urls>); CSP connect-src locked to the two opt-in fetch destinations. - Synthesis Insights = deterministic rule evaluation over your card data. - Chapter Scaffold = string-template generator over your library, grouped by methodology and stance. - Optional GradSummit product-updates sign-up uses a shared Cloudflare Worker; the address you typed is the only thing transmitted. Built by Dr. Rafiq Muhammad, PhD — author of the Mastering Research book series. Part of the GradSummit toolkit for graduate students and researchers.