I apologize for only giving a direct answer to your question. I normally try to add a bit of detail and guidance. But I’m puzzled as to how you lost content in experimenting with links in DEVONthink.
Clickable links can be inserted in rich text notes in DEVONthink in a variety of ways, and the ‘link’ commands are documented in Help and in the user documentation PDF .
One can get the Item Link of any target document (and in the case of PDFs, the Page Link to a specific page) and paste it in as a clickable link in a rich text note.
A text string in a rich text document can be selected. Control-click on the selection and choose the contextual menu option, Link To, then navigate to the desired target.
A text string in a rich text document can be selected. Control-click on the selection and choose the contextual menu option, Add Link, then paste in a URL, Item Link or Page Link.
Control-click on an existing link allows one to choose a contextual menu option, Edit Link, or another option, Remove Link.
Another approach, which also works in plain text notes, is to use well-chosen “cue strings” from a document that one would like to return to, and paste the string (enclosed in quotation marks) into a note. If one then selects that string in the note (including the quotation marks) and presses the keyboard shortcut for the Lookup Service, a new Search window will open with the string entered. Press Return and the desired document will be listed in search results. Select it and (for several filetypes) the document will be displayed, scrolled to the first occurrence of the cue string.
An Annotation note template (with a keyboard shortcut) can be invoked for any selected document. It is already linked to that docent and is associated with it as well by the inclusion of the Name of the annotated item.
And then there are WikiLinks (which I rarely use myself), a feature that can be turned on in Preferences > Editing. Type or paste the Name of any existing document in an open database in a rich text note, and it will be automatically linked to that document.
DEVONthink differs from other database management applications by including artificial intelligence algorithms in the kernel of the database. A your databases grow (especially if they are reasonably well organized) the AI assistants such as Classify (which prompts for possible group location(s) of a new item) and See Also (which suggests other documents that are contextually related to a document you are viewing) become more and more useful.
When I’m doing research and writing in my main database, which holds about 25,000 references and about 5,000 of my own notes, See Also makes DEVONthink the best research assistant I’ve ever had. (There’s a related assistant, See Related Text, that one can use for selected text such as several paragraphs, or a section of a report, for better focus on a concept than if the entire document were examined and compared to all the other documents in a database.)
If I Option-click on a term, I’m presented with a list of all the other documents in that database that contain that term. Or select a phrase and choose the Lookup Service, and a search can be done on all the items in the database that contain that phrase.
Such tools provide a very rich environment for exploring and thinking about the information content of a database. And the best preventative of writer’s block that I’ve ever found.