New: MCP integration. Run plagiarism scans from Claude, Cursor or any AI assistant. Run scans from Claude or Cursor. Set it up

Know every assignment was written by hand.

Window is the free editor students build their assignments in. It records every paste and every minute spent outside the editor while they work, so code that was not typed by the student cannot slip through unseen. When they submit, the project lands in your Codequiry dashboard with that record, and the AI, web and peer plagiarism checks run on top. You end up knowing, not guessing.

  • Every paste is recorded, with its size and when it landed.
  • Every minute off-window is timed, and a paste right after coming back is flagged.
  • Every submission is checked for AI, web and peer plagiarism on top.
4.72/5 · Codequiry is rated by 5K+ verified reviews
solver.py · Window
Monitoring
Explorer
CS201-ASSIGNMENT-4
src
solver.py
graph.py
tests.py
README.md
requirements.txt
solver.py ×
graph.py
src solver.py shortest_path
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
# Dijkstra shortest path over a weighted graph
import heapq
 
def shortest_path(graph, start):
distances = {start: 0}
queue = [(0, start)]
 
while queue:
dist, node = heapq.heappop(queue)
if dist > distances.get(node, float("inf")):
continue
for neighbor, weight in graph[node]:
new = dist + weight
if new < distances.get(neighbor, float("inf")):
distances[neighbor] = new
heapq.heappush(queue, (new, neighbor))
return distances
Integrity
Typed by hand94%
Pastes0
Off-window2m 10s
Authenticity96
Session 1h 12m · record saved with submission
main Monitoring 0 pastes
Class 4F9A2K Ln 15, Col 33 Spaces: 4 UTF-8 Python

Window during a session. Students see the VS Code layout they already know, with an integrity panel beside the editor.

  1. Student opens Window Free download. No account.
  2. Enters your class code Six characters from the assignment.
  3. Writes the project Every paste and tab-away recorded.
  4. Submits from inside Window Lands in your assignment with its record.
  5. Codequiry checks it AI, web and peer plagiarism, side by side.
The editor

A real editor, not a locked-down text box

Window is built on the open-source core of Visual Studio Code, the editor most students already use. It has the file explorer, tabs, syntax highlighting for the usual languages, an integrated terminal, and run and debug. Students do not learn a new tool for your course, and you are not grading code that was typed into a form field.

The difference is a small integrity extension that ships inside it. While a student works, it records the events you would want to see afterwards: text that arrived in one paste instead of being typed, how much of the file was typed by hand, how long the window was out of focus, and whether a paste landed right after the student came back to it. When the student submits, that record travels with the project.

Window checks for updates each time it opens, so the extension is always current. There is nothing for you or your IT department to maintain.

The idea

Detection after the fact tells you something is wrong. It rarely tells you what happened.

AI code detection, web source matching and peer comparison are good at telling you that an assignment deserves a look. They are worse at telling you how the file came to exist. A student who pasted an answer from ChatGPT and a student who retyped it from a second screen hand in the same file, and the conversation that follows is a percentage against a student's word.

Window closes that gap from the other side. The editor itself keeps the record while the work happens, so by the time a project reaches you, the question of whether it was written by hand is already answered. Most students never trigger a single flag. For the ones who do, you are no longer guessing.

How Window gets to "every assignment"

  • Code that was pasted is recorded as a paste, with its size, the moment it landed, and a screen capture of what was on screen.
  • Code that was typed from a second screen shows up as time off-window, a look-and-type rhythm, and then the AI and web checks catch the source.
  • Code that was typed by hand reads as exactly that: a high typed-by-hand share, no pastes, a clear authenticity score.

It also changes behaviour before anything is submitted. When students know the editor keeps a record of how the work was done, most of them do the work.

Setup

How it works

Setup takes a few minutes for you and a few seconds for each student. There is no roster to upload and no student accounts to create.

  1. Create the assignment and open the Window card

    Every assignment in your Codequiry dashboard has a Codequiry Window card on its page. It holds a six-character class code with a copy button. The code uses letters and numbers that are easy to read out loud, so characters like O, 0, I and 1 are never used. Regenerate it any time and the old code stops working.

  2. Share the code and the download link

    Put both in your LMS, syllabus or assignment text. Students install Window on Windows 10 or 11. That is the whole rollout.

  3. Students enter the code and their name

    Window looks up the code and shows the course and assignment title back to them, so they know they are in the right place before they start. They can add a student ID if you ask for one. Then they build the project like any other.

  4. They submit from inside Window

    The project arrives in your assignment as a zip, exactly like an upload, with the integrity record beside it. If the assignment was still a draft, it moves to ready on the first submission. Download everything in one click, or run checks straight away.

Assignment 4 · class code
4F9A2K
Students type this in Window. Nothing else identifies your assignment.
The record

What you see for every submission

The Window results page in your dashboard lists each student on one row. These are the columns, in the order they appear, and what each one means.

ColumnWhat it tells you
AuthenticityA score from 0 to 100 that summarises the whole session. 80 and above reads as clear, 50 to 79 is worth a look, and below 50 is high risk.
RiskClear, Review or High risk, taken from that score so you can sort the class at a glance.
PastesHow many pastes landed in the editor, and separately how many of them arrived right after the student left the window and came back.
Typed by handThe share of the final file the student typed key by key rather than pasted or inserted.
Off-windowHow long the editor was out of focus during the session.
FlagsThe specific events worth your attention, written in plain words rather than codes.
SubmittedWhen the submission arrived.
Assignment 4 · Window resultsCS201 · 28 submissions via Window
Live
StudentAuthenticityRiskPastesTyped by handOff-windowEvidence
Jordan Lee96Clear098%2Record + screens
Priya Shah91Clear193%4Record + screens
Alex Romero31High risk3 +2 after tab-away34%11Record + screens
Sam Carter64Review271%5Record + screens
Dana Wu99Clear0100%0Record + screens

The full record and the screen captures from the session travel in a separate evidence bundle. It is stored privately, apart from the project zip, so students cannot open it and it is never part of the project download you hand to a TA. The Evidence column opens it, one zip per student, with the session's integrity record and every screen capture inside.

Then the usual checks run on top

A Window submission is a normal Codequiry submission. Run AI code detection, web source matching and peer comparison on the assignment the way you always do, and read every signal side by side for each student. Window tells you how the code was written. The checks tell you where it came from. Together they close the gap that a detector on its own leaves open.

Compared

How Window compares with the other ways to handle it

Most programs use one of three approaches today: an honor code, a lockdown browser during exams, or detection after submission. Each covers part of the problem. Window is built for the part none of them cover, which is knowing what happened while the code was written, on ordinary homework.

Honor codeLockdown browserDetection after submissionWindow
Real multi-file projects with a terminalYesLimitedYesYes
Records pastes and typing as the work happensNoNoNoYes
Shows time spent outside the editorNoBlocks it during the examNoYes
Works for week-long homework, not only timed examsYesNoYesYes
AI, web and peer checks on the resultNoNoYesYes
Tells you every assignment was written by handNoNoNoYes
Student needs an accountNoUsuallyNoNo
Cost to studentsNoneOften licensedNoneFree

Where Window fits

University programming courses

Introductory and intermediate courses are where AI-written submissions are most common and hardest to prove. Window gives every assignment a record without changing how students work, and the results sit next to the AI, web and peer checks you already run in Codequiry.

Bootcamps and online programs

Remote cohorts cannot be watched over a shoulder. A class code in the assignment and a download link in the course materials is the entire rollout, and instructors see the record for every student in one table.

Take-home coding interviews

Hiring teams send a take-home and hope the candidate wrote it. Give the candidate a class code instead. You get the project, a record of how it was written, and peer comparison against every other candidate who did the same exercise.

Timed labs and assessments

For in-class work, Window records the session without blocking anything, so it runs alongside whatever supervision you already have. Off-window time and pastes after leaving the editor are the two columns to sort by.

What instructors tell us

Codequiry is rated 4.72 out of 5 across 5,152 verified reviews, and is used at more than 500 institutions. Window is the newest part of it, built by the same team and feeding the same dashboard. These are words from instructors who run Codequiry in their courses.

“Teaching over 500 students per term, we faced a critical challenge: detecting code copied from ChatGPT and online sources without viable tools. Codequiry enables us to batch-process entire cohorts efficiently. This technology is essential for maintaining rigorous standards in computer science education.”

MITSenior Lecturer

“The rise of AI-generated code has made verifying student mastery a critical challenge. Codequiry is now essential for upholding our academic standards and ensuring every submission represents original effort.”

Stanford UniversityProfessor

“Codequiry actually works, detecting AI-generated code and web plagiarism that outdated tools miss. We finally have real enforcement backing our academic integrity policies.”

Georgia TechLecturer

“We found a 99% match between a student and a Chegg solution. I can't believe CS departments are missing this.”

University of TorontoProfessor

A note from Josiah

I built Codequiry because instructors kept telling me the same thing: they could tell when an assignment was not the student's own work, and they could not prove it. Detection got better every year, and the conversation at the end still came down to a number and a denial.

Window is my answer to that. It does not try to lock students out of the internet, because that does not work for homework and it treats everyone like a suspect. It keeps an honest record of how the work was done, and hands it to you with the submission. If you have questions about how it fits your course, write to me. I read every message.

Josiah, founder of Codequiry

What students see, and what is collected

Students see a normal editor with a small integrity panel that says monitoring is on. Nothing is hidden from them. During a session Window records edits and pastes in the editor, when the window loses and regains focus, screen captures, the name and optional student ID they typed in, and the project files they submit. Nothing is collected outside a Window session, and every record is tied to the class code, so it can only reach the assignment that owns it.

Tell students up front. One sentence in the syllabus is enough, and it does most of the work on its own:

Programming assignments in this course are written and submitted in Codequiry Window, which records editing activity and screen captures during the session for integrity review.

Window is covered by the Codequiry privacy policy. Evidence bundles are stored privately and separately from project files, and only the educator who owns the assignment can reach them.

Questions instructors ask

Is Window a real IDE or a locked-down text box?

A real IDE. Window is built on the open-source core of Visual Studio Code, so it has the file explorer, tabs, syntax highlighting, integrated terminal, and run and debug tools students already know. Multi-file projects with folders work the same way they do in VS Code.

Do students need a Codequiry account?

No. Students download Window, type the six-character class code from your assignment and their name, and start working. The only account is yours.

Which operating systems does Window run on?

Window ships for 64-bit Windows 10 and Windows 11 today. It checks for updates each time it opens, so the integrity extension inside it is always the current version.

Does Window block ChatGPT or other tools?

No, it records rather than blocks. A student can still open a browser or an AI assistant. What changes is that the record will show the editor lost focus, and that the code which appeared next arrived in a paste instead of being typed. Recording is what lets Window work for week-long homework, not only for supervised exams.

What exactly does Window collect?

During a session it records edits and pastes in the editor, when the window loses and regains focus, screen captures, the student name and optional ID they enter, and the project files they submit. Nothing is collected outside a Window session, and everything is tied to the class code so it can only reach the assignment that owns it.

Can a student see or change the integrity record?

The record and the screen captures travel in a separate evidence bundle that is stored privately, apart from the project zip. Students never see it, and it is not part of the project download you hand to a grader.

What happens if a class code leaks?

Regenerate it from the Codequiry Window card on the assignment page. The old code stops accepting submissions right away, and submissions already received stay where they are.

How large can a submission be?

Up to 300 MB per project zip. Window packages the project for the student, so there is nothing to compress by hand.

What does Window cost?

Window is free for students. Educators use it with a Codequiry account, and a Window submission counts like any other submission when you run checks on it. Every plan includes it, the free trial too.

Get Window

Educators: create a free Codequiry account, open any assignment, and copy the class code from the Codequiry Window card. Students: download Window, enter the code, and start working.

  • Windows 10 or 11, 64-bit
  • Free for students, no account needed
  • Updates itself on launch
Download Window for Windows Create a free educator account

Already have an account? Sign in and open an assignment to find your class code.