Context Graph
Project Context Graph
Every Academe project has its own context graph: what you’ve read, what you’ve written, what you’re claiming, and what’s still unresolved.
What’s in the graph
Your project graph isn’t just a bibliography. It’s a structured picture of the research process inside a project:
Papers you’ve imported
With the passages you’ve highlighted, the margin notes you’ve written, and the places you’ve already cited them.
Notes and drafts
Your actual writing, the outlines behind it, and the questions you’re still working through.
Claims you’re making
The arguable assertions in your draft, each one linked to the evidence you’re relying on.
Open questions and gaps
Places you flagged “I still need to find out X” - or where Academe noticed something missing.
Data, figures, and results
With the analyses, scripts, and interpretations attached.
Chat and research threads
Exploratory AI conversations kept as part of the project memory.
People and collaborators
Co-authors, advisors, reviewers, and the threads of conversation tied to each.
How it stays accurate
The graph updates as you work. When you write a paragraph, highlight a passage, import a paper, or close out a question, the graph shifts to reflect the current project state.
You can also edit it directly. If Academe misidentifies one of your claims, decides two notes are about the same thing when they aren’t, or flags a gap you’ve already filled in, you can correct it in place and the correction sticks.
What it changes in practice
- Answers use your sources firstChat draws on your papers and notes, not a generic Wikipedia-flavoured summary.
- New writing checked against oldAcademe tells you which existing claims a new section reinforces, contradicts, or depends on.
- Resumable researchRe-open a project after weeks away and pick up the thread: open questions, unfinished drafts, and flagged papers you have not read yet.
- Citations you actually haveCitation suggestions come from your imported library, with the specific passage tied to the claim.
- Export as an organised archiveThe graph becomes a project archive you can hand to a co-author, not just a zip of PDFs.
What stays private
Your project graph is yours. By default it is scoped to the project it belongs to, is not used to train external models, and is not shared with other users. Collaboration within a project is opt-in, per project, per collaborator. Cross-project memory is off by default; if you turn it on, it stays within your account and can be toggled off at any time.
Next