Skip to main content

Tags

Tags in architxt are simple badge-style labels you attach to documents. They work like lightweight categories: Project-Alpha, Billing-Module, Draft, or Reviewed. Tags help you group documents in the UI, scope mental models, and drive Hindsight retention through both tags and document_tags.

A document can have many tags at once. Tags are managed on the Tags page and applied from the Documents page.

Creating and managing tags​

Open the Tags page from the main navigation. The page lists every tag with its Tag ID, Name, Documents count, Generated By source, and creation time.

You can:

  • Click Add to create a tag. Only a name is required.
  • Click Import to paste a CSV of tag names. The optional header is tag_name; one tag per row.
  • Select one or more tags and click Delete. Deleting a tag removes it from any documents that use it; the documents themselves stay in architxt.
  • Search by name or generated-by source.

Click a tag row to open Tag Details. From there you can edit the name. If the tag is assigned to any documents that have already been synced to Hindsight, the dialog warns you that renaming it may cause a mismatch on the next sync.

Applying tags to documents​

Tags are assigned from the Documents page:

  1. Select one or more documents.
  2. Click the Tags button.
  3. The Manage Tags dialog shows:
    • On All Documents — tags currently shared by every selected document.
    • On Some Documents — tags that apply to only a subset.
    • Marked for Removal — tags that will be removed when you save.
    • Available to Add — tags not currently on any selected document.

Click a tag to add it, click it again to remove it, and click Save Changes to apply the result to all selected documents. A document can have many tags at the same time.

You can also assign tags at upload time. In the Upload dialog, choose tags from the shared batch settings and every uploaded document will receive them automatically.

Where tags are used​

Tags serve two roles in Hindsight:

  • They are included in the retain payload as document_tags, so every memory extracted from the document carries those labels.
  • Hindsight can also use tags for tagging memories directly, making them a first-class filter and organization mechanism in the memory bank.

In architxt, tags are also used to scope Mental Models. A mental model's Tags Match setting controls which documents feed the model based on their tags.

How tags fit into architxt​

         Tags
          │
          ├──► Mental Models / Reflect
          │
          ▼
     Documents  ──────►  Hindsight Memory Bank

Tags sit above individual documents. They are not part of the document text, but they travel with the document into Hindsight and influence which documents participate in mental models and Reflect workflows.

Summary​

Tags are quick, flexible labels for documents. Create them once, apply them to many documents, and they become part of both the architxt filtering model and the Hindsight memory-bank tagging model.