The whole projectin one knyte.
Knyte keeps your group’s sources, decisions, questions and next steps in one place, and connects the AI assistants you already use. Everyone works from the same context, so nobody has to explain the project again. You still do the writing itself in your usual docs, slides and sheets.
knyte Swedish, noun a small bundle, tied up in cloth
01One shared understanding
Today, four people means four different AI chatbots and four different understandings of the project.
When everyone copies and pastes into their own AI chat, context drifts apart. Knyte is the one place where the group’s decisions, scope and sources actually hold.
In the bundle
- 01
Purpose and question
What you want to do, the question you want to answer and where the project stands right now.
- 02
Sources
Articles, course literature and the assignment brief. People read the rendered PDF. Assistants read the text that can be extracted, with page references.
- 03
Decisions
Scope and method choices, with the reasoning behind them and the sources that led there.
- 04
Open questions
What you still need to work out together.
- 05
Findings
Claims with evidence and a page reference, reviewed by someone in the group.
- 06
Plan
Tasks, owners and milestones, all the way to submission.
Knyte under your tools.
Knyte = what is true
Sources, decisions, scope, open questions and who does what. Proposed findings stay separate from what you have confirmed. Knyte gives your existing tools and your AI assistants the same context.
Your documents = where you write
Docs, slides and sheets, or whatever you use. This is where you write in your own words.
02Between sessions
Pick up where you left off.
Nobody should have to piece together what happened last time, whether after a week away or before the presentation. In Knyte the context stays, no matter where, when or who.
Without
Where were we again?
With knyte
- MondayAlex
Adds the lecture PDF and the assignment brief.
- WednesdaySofia
Writes down the decision to focus on chapters 3–5, and why.
- ThursdaySofia’s assistant
Reads the sources and proposes a finding. It waits for review.
- MondayNoah
Opens the project before the presentation, after a week away. It is all there.
03Two readers, one project
Your assistant reads the same page.
Connect the AI assistant you already use to the project. It gets the same purpose, the same materials and the same decisions you see, instead of a fresh explanation in every chat.
The connection runs over MCP, the Model Context Protocol, and you approve it yourself.
Project contextActive
Course project
The city we share
- Purpose
- Explain how shared urban spaces shape belonging in a presentation for the course group.
- Current focus
- Reading chapters 3–5. One proposed finding awaits review before the presentation.
- What has been decided
- 01Focus on chapters 3–5
- Materials
- Lecture 4: shared urban spaces.pdfPDF with readable text
- Assignment briefText kept here
- Group presentationLinked only
Written by the people on this project. Assistants can read it; they cannot change it.
Illustrative project
04Visible and shared AI use
AI contributes, but you decide.
Assistants read the context, add notes and tasks, and propose findings. Contributions are labelled as from an assistant and attributed to the member who used it. You own the purpose, decisions and review.
Written by the people on this project. Assistants can read it; they cannot change it.
- Reads
- Purpose, materials, decisions, tasks and notes.
- Contributes
- Notes, tasks and proposed findings. Always marked as the assistant’s.
- You decide
- Purpose and decisions are written by you. Proposed findings are confirmed only after a person reviews them.
Try it
FindingProposed
Informal encounters in shared spaces can strengthen a sense of belonging.
Lecture 4, p. 8
Why not just a shared doc?
- PDFs
- People read the rendered PDF. Assistants get the text that can be extracted, with page references. Findings carry their page reference.
- Structure
- Purpose, findings with evidence, decisions with reasoning, open questions and tasks have a place from day one. You do not have to design a blank canvas.
- Trust
- Assistants propose findings. People review and confirm them. Purpose and decisions stay with the group.
Who it is for
Built for student groups.
A thesis, a course project or a group assignment. Knyte keeps what you know and what needs doing together, so your group and your AI assistants share the same context. It works just as well for a whole course, with lectures, readings and notes in one place. Studying on your own works too: then it is you and your AI assistant sharing the same context.
- Thesis
- Course project
- Group assignment
- Course
Tie it all together.
Bring your sources, your group and whatever you are working out.