Projects & documents.

Organize sources into projects; write documents that cite them. A project is a long-lived workspace for a single research thread — a thesis chapter, a contested fact, a deal, a beat — with its own sources, its own documents, and its own scoped Q&A.

§ 01What is a project

A research run is a single transaction — one question, one report. A project is the persistent context around many runs. It collects sources, holds documents, scopes Q&A, and produces a single exportable artifact at the end.

Projects are the right abstraction when you're working on something that takes more than one query: a literature review, a beat investigation, a competitive landscape, a counterparty file. The library lives at the user level; projects let you carve subsets of that library into focused workspaces.

§ 02Creating one

Create a project from the Projects tab in the workspace, or via the API. A project takes a name, an optional description, and an optional set of starting sources. Once created, you can add sources, write documents, and run scoped Q&A inside it.

Projects are private to your account by default. Sharing controls are part of the workspace surface; see your account settings for current options.

§ 03Adding sources

There are three ways to add sources to a project.

Paste a URL

The fastest path. Paste a URL into the project's source pane and the engine scrapes it, extracts metadata, builds citations, and adds it to the project — no full research run required.

Import from a research run

Select sources from a completed run and import them into the project. Scores, selection rationales, and citations come along.

From the library

Pick from sources you've already saved. Library sources arrive with their full metadata and pre-built citations.

§ 04Documents

Documents are markdown drafts and notes that live inside a project. You can write a literature review, a memo, a brief, or a notes file. Documents support inline citations of project sources — type the citation marker and the project's citation pages handle rendering.

Documents are persistent and editable; they're stored against the project, not the run. You can have many documents per project — one per chapter, one per topic, whatever fits.

§ 05Citation pages

Each project auto-generates a citation page from its sources. The page is rendered in your preferred format and updates as sources are added or removed. When you export the project, the citation page becomes the bibliography of the exported document.

Switching format on a project is one click — every document inside the project re-renders its inline citations against the new format.

§ 06Q&A inside a project

Each project carries a scoped Q&A surface — a chat that is grounded in the project's documents and sources, not the open web. Ask a question and the answer is drawn from material you've explicitly admitted into the project.

Project Q&A is the right fit when you've assembled a corpus and want to interrogate it without expanding scope. For open-web research, run a fresh query on ResearchAnything and import the survivors.

§ 07Exporting a project

A finished project can be exported as a single document — HTML or PDF — with the project's documents combined and the citation page appended at the foot. The exported document carries:

• A title page with the project name and metadata.
• The project's documents in order, with inline citations rendered in the chosen format.
• An aggregated bibliography pulled from the citation page.
• An optional appendix listing the source records with scores and rationales.

The HTML export is self-contained — the citation styles travel with the document. The PDF export is paginated and print-ready.