Right-Rail Module Overview
Choose the right desktop module and move research work into a durable, reviewable state.
The NNScholar desktop app is not designed to keep every action inside one chat box. It separates research work into modules that can preserve, review, and continue the work. When you enter a research space, use this page to decide what you are trying to do, then open the matching right-rail module.
This page is the operating manual for the NNScholar desktop right rail. Use it when you are inside a research space and need to decide which module owns the next action.
The center of the app is the active work surface. The right rail is the set of task modules that help you search, collect, read, write, automate, and inspect evidence without leaving that space. A good workflow does not keep everything in Chat. It moves each piece of work to the module that can preserve it best.
Chat also has a 4-phase academic skill strip above the input: Discovery-Ideation, Planning-Design, Analysis-Experiment, and Writing-Publication. These shortcuts start research tasks quickly; the right-rail modules preserve, read, browse, deliver, and manage the materials produced by those tasks.
Module Map
| Module | Owns | Typical handoff |
|---|---|---|
| Chat | Project reasoning, planning, comparison, synthesis, and next actions | Send stable drafts to Document Writing or artifacts |
| Expert library | Domain experts and reusable evidence-grounded expert skills | Summon an expert into Chat after reviewing its scope |
| Literature | Scholarly search, screening, filters, selected papers, and export | Save useful papers to Library |
| Library | Saved papers, folders, reference sets, PDFs, and reading queues | Open papers in Paper reading or use them in Chat |
| Paper reading | PDF upload, focused reading, translation, evidence extraction, and PDF Q&A | Move verified PDF evidence into Chat or Document Writing |
| Document Writing | Manuscript drafting, revision, format matching, document Q&A, and review | Export, review in Deliverables, or prepare submission |
| Deliverables | Generated files, previews, provenance, and completed research assets | Open, reveal, download, or trace back to the source context |
| Research automation | Repeating research tasks, topic alerts, paper recommendations | Review automation outputs in the target workspace |
| Plugins and skills | Installed apps, plugins, skills, MCP capabilities, and extensions | Enable capabilities before asking Chat or experts to use it |
| Browser | Publisher pages, external databases, URLs, and source inspection | Save useful evidence back to Literature, Library, or Chat |
| Side chat | Focused side reasoning from the current workspace snapshot | Fold useful decisions back into the main thread |
Each module page includes a screenshot from the current desktop interface, so users can recognize what the module looks like when it opens and where the first useful action usually starts.
How To Use This Section
- Start with the module map above and decide whether the current task is search, reading, writing, deliverable review, automation, source inspection, or capability setup.
- Open the matching page from the left navigation. Every module page follows the same pattern: when to open it, what the screenshot shows, what to use it for, how to proceed, and what to check before leaving.
- Hand the result to the next module when the step is complete. Useful papers, PDF evidence, manuscript revisions, and generated files should not remain only in a temporary chat.
Common Decisions
- Unsure what to do next: open Chat.
- Need to find papers: open Literature.
- Need to organize selected papers: open Library.
- Need to read, translate, or verify a PDF: open Paper reading.
- Need to draft or revise a manuscript, match formatting, ask questions about a document, or run document review: open Document Writing.
- Need to inspect generated files: open Deliverables.
- Need a recurring research routine: open Research automation.
- Need plugins, skills, or MCP capabilities: open Plugins and skills.
- Need to inspect a web source or DOI page: open Browser.
- Need to prepare a journal submission portal workflow: choose Writing-Publication -> Automatic submission assistant in Chat, then use Browser for the journal portal.
- Need a temporary focused question beside the main thread: open Side chat.
Example Demo
Scenario: you start from a new research idea: "Can PCAT on coronary CTA predict cardiovascular events in patients with diabetes?"
Run the module chain like this:
- Chat: break the idea into search, reading, study design, analysis, and writing plans.
- Literature: search for papers about PCAT, CTA, diabetes, and prognosis.
- Library: save core papers and organize them into
core-references,methods, andbackground. - Paper reading: open core PDFs, ask PDF-grounded questions, and save page-level evidence.
- Chat skill phases: choose Planning-Design for the study plan, then Analysis-Experiment for the statistical plan.
- Document Writing: turn the study design and evidence into a manuscript draft, then run format matching and document review.
- Browser + Automatic submission assistant: open the journal portal and check fields and upload requirements.
- Deliverables: save manuscript versions, figures, evidence tables, and submission materials.
This example shows that the right rail is not a set of isolated buttons. It is the pipeline that moves a research task from idea to submission-ready materials.
What Counts As Complete
A right-rail workflow is complete when the output is in the module that should preserve it, the evidence can be traced back to papers or files, and the next action is obvious. If the result only exists as an unsourced chat answer, it is not complete yet.