Graph View

The graph view renders your knowledge graph as an interactive network visualisation. You can explore connections between entities, filter by type and time period, and save custom views.

Access points

• World graph: Shows all entities and relationships in the world. Access via Graph in the world sidebar.
• Entity graph: Centres on a single entity and shows its neighbourhood. Access from an entity’s detail page.

Layout algorithms

Four layout options are available in the toolbar:

Layout	Best for	Description
Force-directed	Exploring complex networks	Physics simulation with attraction/repulsion. Default for all graphs.
Hierarchical	Showing parent-child or rank structures	DAG-based with top-down or left-right flow. Minimises edge crossing.
Circular	Equal emphasis on all nodes	Nodes arranged in concentric rings.
Concentric	Entity-focused exploration	Centre entity in the middle, surrounding rings by graph distance.

Depth control

In entity-centred graphs, the depth control sets how many hops from the centre entity to display (1–5). At depth 1, you see only directly connected entities. At depth 5, you see entities up to 5 relationships away.

Note

Depth controls are only visible in entity-centred graphs. World graphs display all entities matching the current filters.

Node limit

Big worlds — an imported Azgaar map can hold close to a thousand entities — would grind the layout to a halt if everything rendered at once. The Max selector in the toolbar caps how many nodes load (100 up to 3,000; the default is 500, a comfortable ceiling for most machines). When your world holds more than the cap, a banner tells you the view is truncated so you know you’re seeing a sample, not the whole world — raise the limit, or narrow the filters to bring what you care about into view.

Status line

The corner of the canvas keeps a running count of what’s on screen: 128 nodes · 214 edges. The moment a filter is hiding something — a type unticked in the rail, a field constraint, a narrowed temporal window — a third figure joins it: 128 nodes · 214 edges · 1,076 hidden by filters. That figure only shows up while a filter is actually holding something back, so a plain two-part count is your reassurance that nothing in your current filters is keeping the rest of the world out of view. It answers the question a hard filter provokes at a glance: has the world gone missing, or have you just narrowed to a corner of it?

Temporal controls

The Time button in the toolbar (the clock) asks which moment you want to see. Live — the world as it stands is the default. The other three choices each pick a moment and ask a question about it:

Choice	Shows
At a moment · Alive then	Entities that existed at the moment you pick
At a moment · Not yet	Entities that hadn’t come into being yet
At a moment · Already gone	Entities that had ended by then

Choosing any of them opens the temporal bar along the bottom of the canvas, where you set the moment:

The moment starts on your world’s first date, the earliest date anything in it carries, so in Verenthia it opens on Mythara’s Bloom 17, Year 1 rather than on an empty canvas before anything existed. Drag the scrubber forward from there.

Close the bar with its × to go back to Live.

Point-in-time mode

Shows entities and relationships that were active at a specific moment. Useful for answering “what did the world look like on this date?”

Range mode

Shows entities and relationships active during a time span. Useful for seeing everything that was relevant during a war, a reign, or a season.

Playback

The temporal bar includes playback controls that step through time:

• Play/Pause — auto-advance through time
• Speed — 0.5x, 1x, 2x, 4x
• Step forward/backward — move by one of your calendar’s own units. The step menu lists them, divisions of the day included, so in Verenthia you can step by Year, Month, Day or a single Bell. It starts on your calendar’s day.
• Timeline scrubber — drag to jump to a specific point. A moment within a day reads with its time, Mythara’s Bloom 17, Year 1 at Bell 1

Entity type filters

Down the left side of the graph sits a column listing every entity type in your world, each one coloured to match its nodes — it’s the same column that doubles as your colour key, so ticking a type here is also how you’d look up what a colour means.

Types are grouped the way you’ve grouped them in Manage → Entity Types, in the order you put them in — see Organising Your Entity Types. A type you haven’t filed into one of your groups sits under Custom Types. On the rare occasion your groups can’t be loaded, the column simply lists every type alphabetically instead, so you’re never left without it.

A search box at the top of the column finds a type by name — handy once a world has a couple of dozen of them. It also matches the short id a type is stored under, so typing magic_system finds Magic System. Relationship types are searched by name.

Note

Searching only changes what the column shows. It never moves the graph, and a type you’ve checked stays checked and stays applied to the graph even while a search has scrolled it out of view.

Each type shows as a checkbox — uncheck types you want to hide. Nothing checked shows every type, and the column says so while that’s the case. When you filter entity types, the available relationship type filters update automatically to show only types that connect the visible entity types.

A type carrying field conditions shows the number of conditions beside its entity count, so you can see at a glance which types are narrowed without opening the filter bar below — see Field-based filters for how to add one.

A type with nothing loaded in the current graph is greyed out and can’t be checked, because checking it could only ever produce an empty canvas.

Relationship type filters

Toggle which relationship types are visible. The search box at the top of the column narrows this list too, by name. Filtered by the currently selected entity types — only types that can connect visible entities are shown.

Kinds of relationship

Many relationship types hold several kinds. Member of covers a guild’s Leader, its Officers and its rank-and-file Members; Created covers things born, built and crafted. To narrow a type down to some of its kinds, click the arrow beside it in the column. The list shows every kind its links on the graph carry, each with a count, and Unspecified for links that were never given a kind.

Untick a kind to hide just those links; the rest of the type stays. Unticking Leader under Member of leaves the Officers and Members in place, and unticking Unspecified hides only the links nobody gave a kind. While any kind of a type is hidden, the type’s row says how many are showing, such as member of (19 of 20 kinds).

Kinds and the type checkboxes work together: a type you’ve left out hides all of its kinds, and a kind you’ve hidden stays hidden when its type is the one you’re looking at. Clear at the top of the section brings back every type and every kind.

Kinds are matched regardless of capitals, so Leader and leader (which can happen when links come from an import) count as one kind.

The kind list lives in the column, so it’s on screens wide enough to show it. On a narrower screen the toolbar’s relationship filter works by type only, though a saved view or a shared link with hidden kinds still applies them.

Mentions

The Mentions toggle in the toolbar (icon: 📄) overlays the content units (scenes, articles, sessions — whatever your content types are) that your visible entities are mentioned in.

When enabled, each mentioned content unit appears as a distinct dashed, rounded-rectangle node (clearly different from the circular entity nodes), connected to the entities it mentions by a mentioned in edge. This lets you see, at a glance, which documents reference a character, location, or item — and which entities tend to appear together in the same scenes.

Click a content-unit node to open its details in the sidebar (title, narrative importance, word count, summary) with an Open in Editor button that jumps straight to that document.

Note

Mention links are created when you run entity detection on your content (or when AI extraction processes it). A world where entities were only linked manually in the editor may not show mention edges until detection has run over its content. Off by default.

Reverse edges

The Reverse toggle in the toolbar (icon: ↔) controls how asymmetric relationships render.

For an asymmetric relationship like A is a member of B, MythTapestry stores two edges in the graph database:

• A primary edge from A to B with the type MEMBER_OF
• An auto-created reverse edge from B to A with the type HAS_MEMBER

This lets each side describe the relationship in its own language (“I’m a member of B” vs “I have A as a member”) without you having to author both directions.

Off (default): Typically only the primary edge renders, so you see one arrow per logical relationship and the graph’s edge count agrees with the entity’s relationships list. (Legacy edges authored before the dual-edge architecture pre-date the is_primary flag and remain visible — if your world predates this change, you may still see both directions for some relationships.)

On: Both edges render. You’ll see two arrows between the connected entities — one labeled with the forward name (Member of) and one with the reverse name (Has member). Useful when you want both perspectives visible at the same time, or when debugging unexpected results.

The same toggle exists on the entity detail → Relationships tab (button labeled “Reverse edges”). It controls the relationship list and the embedded graph view together. Both views agree within whichever mode you’ve selected.

Symmetric relationships (e.g., Allied with, Married to) only ever have one edge; the toggle has no visible effect on them.

Field-based filters

For advanced filtering, add field constraints that filter entities or relationships by their property values. Field constraints appear as a chip bar below the main controls — click a chip to edit its operator or value.

Adding a constraint

1. Click the + button on an entity or relationship type chip to open the constraint editor.
2. Choose a field from the dropdown (e.g. name, status, a custom attribute).
3. Choose an operator (see table below).
4. Enter a value in the text input and press Enter to confirm.

Operators

Operator	Applies to	Example
equals	Text, numbers	status equals “active”
not equals (≠)	Text, numbers	status ≠ “inactive”
contains	Text	name contains “council”
not contains	Text	name not contains “draft”
starts with	Text	name starts with “the”
ends with	Text	title ends with “lord”
regex	Text	name regex “^Lord\s”
= / ≠	Numbers	age = 25
> / ≥	Numbers	age > 100
< / ≤	Numbers	level < 5
between	Numbers	level between 1 and 10
in	Enum / select fields	class in “warrior; mage; rogue”
not in	Enum / select fields	class not in “civilian”
related to	Reference fields	mentor related to “Seraphine”
not related to	Reference fields	faction not related to “Empire”
is empty / is not empty	All types	description is not empty
is true / is false	Boolean fields	is_active is true

Case sensitivity

Text operators (contains, equals, starts with, ends with, and their negations) are case-insensitive by default. Typing v will match “Valentina”, “village”, and “VAULT”.

Tip

Case-insensitive matching applies to all text operators. There is no exact-case mode — all comparisons normalise both sides to lowercase before comparing.

OR alternatives within a filter (semicolons)

To show entities matching any one of several values, enter your terms separated by semicolons in the same value field:

• Type valentina; grand inquisitor in a name contains constraint → matches entities whose name contains “valentina” or “grand inquisitor”.
• Press Enter, Tab, or type a semicolon to commit each term as a separate chip inside the input.
• Click the × on any chip to remove that term.
• Each term is independently case-insensitive.

Note

Semicolon-separated OR works for text operators only (contains, equals, starts with, ends with, and their negations). Numeric, boolean, and date operators always take a single value.

Combining multiple constraints

When a type filter has more than one field constraint, the field logic setting (AND / OR, shown in the toolbar) controls how they are combined:

• AND (default) — an entity must satisfy all constraints to appear.
• OR — an entity appears if it satisfies any constraint.

This combines with the semicolon OR within a single constraint: name contains "valentina; grand inq" is an OR within that one constraint, applied regardless of the global AND/OR setting.

Appearance

Open Graph Settings and choose the Appearance tab to change how a type of entity looks on the graph. Everything you change shows straight away — on the graph behind the panel and on the preview at the top of the tab, a single node drawn exactly the way the graph draws that type. The preview is there for when the panel covers the part of the graph you care about.

Pick the type with Entity type, then:

• Colour — Type default is the type’s own colour. The six swatches beside it are the palette every type is drawn from, so you can give Factions the colour your Organizations already use, say. These swatches follow your theme: switch from a dark theme to a light one and a node you painted with the green swatch stays the light theme’s green, not a washed-out copy of the dark one. The last swatch, Custom (the +), takes any colour as a hex code, like #4A90E2.
• Size by — Connections makes the most connected Characters the biggest; it’s what every type does until you say otherwise. A field sizes by a number you track, such as a Character’s age or a Settlement’s population. Fixed draws every node of the type at one size.
• Label — show or hide the name, put it inside, above, below, left or right of the node, and set its size.
• Show type icon — the type’s icon inside the node.
• Reset Character to default puts that one type’s appearance back the way it started. It leaves your visual rules alone.

Note

Size by is a visual rule underneath. Choosing A field or Fixed adds a size rule for the type, and you’ll find it — with its smallest and largest sizes and its scale — under Advanced. Trying Fixed doesn’t throw away a rule you’ve tuned: it waits underneath, and picking that field again brings it back as you left it. If a type has more than one size rule, Size by shows the first and tells you to look in Advanced for the rest.

Advanced ›, at the bottom of the tab, holds everything else: visual rules, which field the label shows and its colour, border width, where the icon sits and how big it is, what the hover popup lists, edge actions, and the canvas. ‹ Appearance takes you back with the same type still selected.

Canvas

The canvas is the background behind the graph. Theme default, Deeper and Softer each come from your theme, so they suit whichever theme you’re using. The choice applies to every graph you open. If you picked a plain colour such as white with the old settings, it keeps showing — as Custom — until you choose one of the three.

Visual rules

Visual rules size or fade the nodes of one type by a number. They live under Graph Settings → Appearance → Advanced, and each one reads as a sentence: Characters sized by Population, Locations faded by connections.

• Click a rule to tune it. A size rule has a smallest and largest size in pixels; an opacity rule has a lowest and highest opacity between 0 and 1.
• Scale decides how values spread across that range: Linear for evenly spread numbers, Logarithmic when a few values are far bigger than the rest, Square root in between.
• When the field is empty draws entities that have no value at the default size, or hides them.
• The switch beside a rule turns it off without deleting it; the × removes it.

Add a rule with the form underneath the list: choose Size or Opacity, the type, then the field. Connections — how many relationships an entity has — is always on offer, even for types with no number fields. While any rule is on, the Appearance tab carries an Active badge.

Tip

If a rule’s field has been renamed or removed from the type, the rule says so — Field age no longer exists — and the Appearance tab points you to it. Remove the rule, or add a new one for the renamed field.

Label colour

Each node’s name can be tinted independently of the node itself, from Graph Settings → Appearance → Advanced → Label details.

Left on Auto (the default), names follow your theme: readable against the bare canvas when the name sits outside the node — above, below, left, or right of it — and switched to white when the name sits Inside, where the node’s own colour is right behind it. Choose Custom and pick a colour, and the name is drawn in that colour wherever it’s positioned.

Note

If you’d set a colour before and never actually saw it on the canvas — the swatch changed but every name still drew white — that’s fixed. Your setting now reads as Auto until you pick a colour again.

Saved views

Save your current graph configuration — filters (including any hidden kinds of relationship), depth, layout, temporal mode — as a named view that you can restore later.

• Create: Save current state with a name and optional description
• Apply: Restore all settings from a saved view
• Update: Overwrite a saved view with current settings
• Set as default: Automatically load this view when opening the graph
• Delete / Rename: Manage saved views

Each view in the list carries a coloured dot and a line describing what it actually does — Characters + Organizations · depth 3, or All types for a view that narrows nothing. The name is what you called it; the line underneath is what you’ll get, so you can pick the right one without applying three of them to find out. It mentions the types the view filters to, how many conditions narrow them, how many kinds of relationship it hides, the depth when it isn’t the usual two, and whether the view is pinned to a moment or a stretch of time.

The dot is just a marker to find the same view by twice — it’s picked for you and stays put.

Saved views are stored per world, so different worlds can have different default views.

Compact mode

Toggle compact mode to reduce the toolbar to a single condensed line. This gives more screen space to the graph while keeping all controls accessible via dropdowns. Compact mode is enabled by default.

Fullscreen mode

The Fullscreen button in the graph header expands the graph to fill the entire window, hiding the app’s top navigation and the world sidebar. The filter toolbar stays in place, and the details sidebar still opens when you select a node or edge — so you keep every control while gaining the maximum amount of canvas.

Available on both the world graph and the entity-centred graph. Press Esc or click Exit to return to the normal layout.

Interacting with the graph

Mouse controls

Action	Effect
Click node	Select node — it grows and the links running to it brighten, and the details sidebar opens (content-unit nodes show document details with an Open in Editor button)
Click edge	Select edge and show relationship details
Double-click node	Centre and zoom in on that node
Ctrl/Cmd + click node	Navigate directly to entity detail page (no effect on content-unit nodes)
Right-click	Context menu (create entity, create relationship, focus, delete)
Drag node	Reposition it in the layout
Box select	Click and drag on empty space to select multiple nodes
Scroll wheel	Zoom in/out
Fit to view button (or 0)	Zoom out until everything shown fits on the canvas. A world of a few hundred entities can need 10% zoom or less to fit; with a filter on, it fits only what the filter leaves
Click + drag background	Pan the view

Tip

A selected node grows relative to whatever size it already had, so a size rule mapping connections to node size still holds — a hub that was already the biggest node on the canvas grows from that larger base rather than jumping to one fixed size. The highlight on its links stays put as you move the mouse elsewhere on the canvas, so you can see what a node touches without losing your place.

Keyboard shortcuts

Shortcut	Action
Ctrl/Cmd + E	Toggle edge actions mode (hover nodes to see quick-create actions)
Ctrl/Cmd + Shift + E	Open bulk entity creation modal
Esc	Exit fullscreen mode

Edge actions mode

When edge actions are enabled (via the toolbar toggle or Ctrl+E), hovering over a node reveals action buttons on its edges. Click these to quickly create relationships between entities without opening a separate dialog.

Tip

Edge actions mode disables node dragging. Toggle it off with Ctrl+E when you want to reposition nodes.