ThinkForge uses tags to organize and find documents across a project. Tags are stored alongside the document's indexed content, not in sidecar files on disk. Every indexed document can have tags added, removed, and searched.
Adding and Removing Tags
Tags can be managed from several surfaces:
Document Info Panel — The right sidebar in Project Navigator shows editable tag chips for the selected document. Each tag has a remove button. An "Add tag..." text box with an Add button lets you create new tags. Pressing Enter submits the tag.
Rich text editor — The editor includes tag management in its toolbar area.
Files — Document cards support tag addition.
Forge Graph node editor — The node detail panel shows your tags (editable) and system tags (read-only chips) separately.
Adding a tag to a file that has not yet been indexed triggers indexing first.
User Tags vs System Tags
Tags fall into two categories.
User tags are free-form strings you create. They have no required format and can be anything meaningful to the project — topic labels, status indicators, priority markers, or any other scheme. User tags are the ones shown in the folder tree's Tags section, the Overview tags card, and tag search results.
System tags use reserved prefix:value formats and are generated automatically. They cannot be created manually and are filtered out of user-facing tag lists. System tag prefixes include:
category:— Document category assigned by the project wizard, for examplecategory:System Designwizard:— Marks documents created by the wizard, for examplewizard:generatedtemplate:— Identifies the template a document was generated fromproject:— Project type metadata, for exampleproject:type:App Developmenttech:— Technology tags from the wizard, for exampletech:ReactnodeLevel:— Mind map hierarchy levelmindmap:— Mind map node linksimport:— Import source tracking, for exampleimport:source:chrome_extensionagent:— Agent-generated content markerstool:— Tool classification for the agent tools panel
Wizard-created documents automatically receive system tags based on the template, category, and technology selections.
Tag Search and Filtering
Tags drive filtering and discovery across several surfaces:
Folder tree Tags section — Shows a hierarchical list of your tags. Clicking a tag filters Files to documents carrying it.
Overview — The Tags card shows tags sorted by document count. Clicking a tag card expands that tag's documents in the info panel.
Files tag filter — Switches Files to tag-filtered mode, querying the index for matching documents instead of enumerating the filesystem.
Folder tree Search tab — The tag search mode performs exact, case-insensitive tag matching. Tag chip suggestions appear below the search box.
MCP tag search — Connected AI clients can search by tag with AND logic (all required tags must be present) or OR logic (any matching tag qualifies).
Cluster View — The Forge Graph can group documents by tag, creating visual clusters based on tag membership.
Document Metadata
Beyond tags, ThinkForge tracks metadata for every indexed document:
File information — Absolute path, project-relative path, file type, MIME type, file size, creation date, and last modified date.
Content — The full extracted plain text of the document, used for search and preview.
Content hash — Used for change detection during re-indexing and for detecting file moves, where the same content appears at a different path.
Indexing state — Whether the document is indexed, when it was indexed, and any error from the last indexing attempt.
Sync timestamps — Separate timestamps for file sync, metadata sync, and Chrome extension sync.
Extended metadata — Additional properties such as a display title, for documents whose display name differs from the filename.
Document metadata is visible in the Details panel when a document is selected, and in the focused document view in Files.
Project-Level Tags
Projects themselves can have tags, stored in project.json separately from document tags. Project tags describe the project as a whole rather than individual documents within it.