Moodle owns the context
Course roles, assignment settings, student access, and grading policy remain inside the institution's LMS.
Integrations / Moodle
Moodle AI detection built around educator review
Connect assignment text to the AI Detector API, return sentence-level signals to an institution-controlled workflow, and keep academic decisions with educators. Each deployment starts with a clear review policy and a technical scope.
Moodle deployments are scoped with the institution before implementation.
Review flow
From assignment text to an educator decision
- 1
Moodle
A permitted submission enters review
Your LMS rules decide which text may be processed and who can start a check.
- 2
Connector
The integration extracts and sends text
A server-side request keeps the API key outside the browser and student view.
- 3
Educator
The signal returns for contextual review
A person reads the result beside drafts, sources, and course policy.
01
API-based connection for Moodle workflows
02
Server-side API key authentication
03
Sentence-level detector output
04
Human review remains the final step
A practical boundary
Let Moodle manage learning, and let the API handle analysis
A sound integration does not move every LMS concern into the detector. Moodle keeps users, roles, assignments, permissions, and course context. The connector sends approved text to the API and returns only the evidence an educator needs to review.
The connector protects the key
API credentials belong in a trusted server environment. They should never appear in browser code or student-facing pages.
The detector returns a signal
Use the structured result to focus a review, then consider the writing history and the assignment context.
What the connection needs
Four decisions to make before anyone writes code
These choices shape the connector, the user experience, and the institution's operational responsibility.
Eligible submissions
Define which assignment types are in scope and how supported files become plain text before analysis.
Authorized reviewers
Choose who can run a check, see sentence details, and record any follow-up in the course workflow.
Result placement
Decide whether results belong in the grading view, a separate queue, or a staff-only report.
Review and retention policy
Document how signals may be used, which records are kept, and what appeal or second-review process applies.
Setup path
Start with one course and a written review policy
A narrow pilot makes permissions, extraction errors, result language, and staff training easier to evaluate before wider use.
- 1
Map the assignment workflow
List the submission types, Moodle version, roles, review points, expected volume, and the exact text that may leave the LMS.
- 2
Build and test the connector
Keep the API key server-side, request detector version v3 for new work, and test realistic content plus empty, unsupported, and failed submissions.
- 3
Run a monitored pilot
Train reviewers on the limits of detection, collect false-positive reports, and adjust the workflow before adding more courses.
Good fits
Useful when a review process already exists
The integration works best when the institution already knows who reviews unusual writing signals and what evidence belongs in that conversation.
This is not a misconduct engine
Do not auto-grade, reject, or accuse a student from a detector result. Short, translated, formulaic, or heavily edited writing can produce misleading signals.
Assignment triage
Add a detector signal to the material an educator reviews without turning the score into an automatic penalty.
Course pilots
Test the workflow with a small group of trained instructors before making it available across departments.
Writing conversations
Use highlighted passages to ask focused questions about drafting, sources, and revision choices.
Staff review queues
Route selected submissions to an institution-managed queue when a second reader is part of the policy.
Moodle FAQ
Questions to settle before a pilot
Confirm the technical path, governance, supported content, and review responsibility before presenting the integration to instructors.
Is there a public Moodle plugin I can install myself?+
We do not list a public self-install plugin on this page. Moodle connections are scoped with each institution and use the AI Detector API. Contact the team to confirm the deployment path, supported Moodle version, and implementation responsibility before planning a launch.
What does the Moodle integration send to the detector?+
The detector API accepts plain text. A Moodle-side connector must extract only the assignment text that your institution has chosen to process, then send that text with the selected detector version.
Can it review uploaded PDF or DOCX assignments?+
That depends on the connector. The detector API receives text, not uploaded files, so file extraction must happen before the API request. Confirm the file types and failure handling during scoping.
Does a high detector score prove academic misconduct?+
No. AI detection is a probabilistic signal. Educators should review the writing, drafts, citations, revision history, assignment conditions, and the institution's policy before reaching a conclusion.
Where will instructors see the result?+
Result placement is an implementation choice. It may appear in a grading workflow, a separate review queue, or an institution-controlled report. Agree on roles, visibility, and record keeping before development begins.
Does the integration require an API plan?+
Yes. Detector requests use an API key and consume the detector word balance attached to the active API subscription. Estimate usage with representative assignments before a broad rollout.
Plan the integration
Bring your Moodle version, workflow, and review policy
Tell us which assignments you want to review, who should see the result, and how much text you expect to process. We can then confirm whether an API-based connection fits.