Thogits
A thogit is an entry — an ID, a name, and an optional description. On its own it carries no structured data. Structure comes from tags.
Architecture
Thogits has three core concepts:
Tags
Tags are not labels — they are schemas. Each tag defines a set of typed fields (String, Number, Date, Select, Reference, etc.). Applying a tag to a thogit gives it those fields.
Field Values
When a tag is applied, its field values are stored as JSON keyed by tag ID. Each tag’s data also tracks the schema version at the time it was applied.
Data Model
Section titled “Data Model”┌──────────────────────────────────────────────────────────────┐│ Thogit ││ ├── id: ULID ││ ├── name: String ││ ├── description: String (optional) ││ └── data: Map<TagId, TagData> ││ └── TagData ││ ├── field_values: Map<String, JSON> ││ └── schema_version: u64 │├──────────────────────────────────────────────────────────────┤│ Tag ││ ├── tag_id: ULID ││ ├── name: String (unique) ││ ├── description: String (optional) ││ ├── fields: Map<String, Schema> ││ └── schema_version: u64 │└──────────────────────────────────────────────────────────────┘How Composition Works
Section titled “How Composition Works”A single thogit can have any combination of tags. There is no fixed type hierarchy.
{ "id": "01JQX...", "name": "Fix login timeout bug", "description": "Session expires after 5 minutes instead of 30", "data": { "01JQA...": { "field_values": { "status": { "variant": "InProgress" }, "priority": ["High"] }, "schema_version": 3 }, "01JQB...": { "field_values": { "assignee": "01JQC...", "due_date": "2025-04-15" }, "schema_version": 1 } }}Here one thogit has two tags applied — perhaps “Task” and “Assignment” — each contributing their own fields. Remove a tag, and its fields disappear. Add a new tag, and the thogit gains new fields.
Schema Versioning
Section titled “Schema Versioning”Tags track a schema_version that increments whenever their fields change. When a tag is applied to a thogit, the TagData records the version at that point. This makes it possible to detect when a thogit’s tag data was written against an older schema.
Why This Matters
Section titled “Why This Matters”Traditional knowledge tools force you to choose a structure upfront. Thogits inverts that:
- No rigid types. A thogit is whatever its tags say it is.
- Composable metadata. Need a “Book” that is also a “Project”? Apply both tags.
- Schema evolution. Change a tag’s fields without migrating existing data. Versioning tracks the drift.
- Powerful filtering. The flat field-value storage enables a composable filter DSL that queries across any tag’s fields.