How it works¶

The connector talks to Moodle through its Web Services REST API (/webservice/rest/server.php) using a web-service token. Everything it can see is what that token's user can see.

The walk¶

Courses — it lists the courses the token's user is enrolled in (core_course_get_courses).
Sections & modules — for each course it reads the course contents (core_course_get_contents): the sections (topics/weeks) and the activities (modules) in them.
Documents — each supported module becomes one RAGflow document. Unsupported module types are skipped.

A RAGflow knowledge base with documents synced from Moodle courses Synced Moodle items appear in the knowledge base, each labelled Course / Section / Name.

What becomes a document¶

Moodle module	How it is indexed	Document id
Resource (a file)	The file is downloaded (the token is appended to the file URL) and indexed as-is, keeping its type (PDF, DOCX, PPTX, …).	`moodle_resource_<id>`
Page	The page's stored HTML is downloaded and indexed.	`moodle_page_<id>`
Book	Each chapter's `index.html` is fetched and converted to Markdown, then concatenated into one document.	`moodle_book_<id>`
Forum	All discussions are fetched (`mod_forum_get_forum_discussions`) and rendered to Markdown (one heading per discussion).	`moodle_forum_<id>`
Assignment / Quiz	The activity's description is converted to Markdown (the submissions/questions themselves are not read).	`moodle_assign_<id>` / `moodle_quiz_<id>`

Skipped: label and url modules, and any module type not in the table above. A resource with no file, or a page/book with no content, is skipped too.

Each document gets a human-readable semantic identifier of the form Course fullname / Section name / File-or-activity name, so a citation shows where in Moodle it came from.

Stable ids, so updates land cleanly¶

Document ids are derived from the Moodle module id (moodle_<type>_<id>), not from the content. When a course item changes and is re-synced, it replaces the existing document rather than creating a duplicate; when an item is deleted in Moodle, its document is removed from the index (see Sync behaviour).

Access & downloads¶

Protected Moodle files are served behind authentication, so the connector appends token=<token> to each file URL when downloading. The token therefore needs the capability to download course files in addition to reading course contents — see Setup.

What it does not do¶

It does not read student submissions, quiz questions/answers, grades or messages — only the content listed above (for activities, only their description).
It only sees enrolled courses of the token user; it is not a site-wide crawl unless that user is enrolled everywhere (see Security & permissions).
It does not set an external_sharing flag — see Using it with the Moodle RAGflow Suite.

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.