Security & permissions¶

The token defines the scope¶

The connector reads Moodle as the token's user. It can see exactly what that user can see — the user's enrolled courses and the material in them — and nothing else. Choosing the right account is therefore the main access-control decision:

Use a dedicated service account, not a real person's login. Enrol it (e.g. as a non-editing teacher) in exactly the courses whose content should be searchable, and no others.
Removing the account from a course, or disabling the account, stops new syncs and removes its documents on the next run.
The account needs to read course contents and download course files; it does not need editing or admin rights.

Treat the token like a password¶

A web-service token is a long-lived credential that grants that user's access. Store it only in the RAGflow data source; rotate it if it may have leaked (Manage tokens → delete and re-create), and the data source keeps working once you paste the new one.

What data leaves Moodle¶

The connector copies the content it indexes (files, page/book/forum text, activity descriptions) and its metadata (course, section, module, timestamps, and — for forums — discussion authors) into the RAGflow knowledge base. Consider this when indexing courses with personal data:

Forum documents include discussion titles and the author name/id of each discussion.
Files are copied verbatim into RAGflow; anything inside them (including personal data) goes too.
It does not copy student submissions, grades, quiz responses or private messages.

Once in RAGflow, the material is subject to RAGflow's own access controls and to whatever assistants or plugins are pointed at that knowledge base — so scope the knowledge base and its consumers accordingly.

Data protection checklist¶

Index only courses you are cleared to make searchable.
Prefer a dedicated, least-privilege service account.
Be deliberate about forums (author names) and files (embedded personal data).
Restrict who can query the resulting knowledge base in RAGflow.

The connector is open-source and part of RAGflow; its behaviour can be audited in the RAGflow repository.

Note: This documentation was created with the help of AI. Spotted an error, an omission, or something unclear? Please report it in the ragcon-docs GitHub repository so we can fix it.