NotaeLibrary
The timeline is the epistemic backbone of any collaborative inquiry. Unlike traditional timeline tools that assume events have a single known time, NotaeLibrary embraces uncertainty as a first-class concept — time ranges, competing scenarios, subject layers, and AI-powered analysis all work together to help your team reconstruct events from fragmentary evidence.
On this page
Timeline entities are a type of World entity— alongside Characters, Places, Objects, and Context. They represent events that occurred (or may have occurred) at specific points in time. Each timeline entity lives in your room's bookshelf and has its own discussion thread, task linking, and graph integration.
Timeline entities appear in the World → Events section of your study room. From there you can view, reorder, edit, and analyze them — either as a flat list or as a chronological timeline with range bands.
Each timeline entity carries:
There are two ways to add timeline events to your room:
Create manually in the World section
Navigate to World → Events and click the + Add Event button. Fill in the title, description, dates, tags, and subject. This is best for events you already know about and want to document directly.
Import from a World
Create a World at Home Room → Workspace → Worlds with your entities (characters, places, timeline events). Then attach that World to your study room. The entities are imported as bookshelf items — including their dates, which now survive the import. This is ideal for pre-built mystery scenarios or research chronologies.
Extract from uploaded documents
Upload documents (PDFs, text, web articles) to the bookshelf. The platform's entity extraction pipeline identifies timeline events, characters, and places automatically.
Not every event has a precise timestamp. A witness might say "sometime between Tuesday and Thursday," or a historical record might date an event to "circa 1850." The timeline handles this through range bands.
Range Start
The earliest possible time for this event.
Range End
The latest possible time for this event.
Display Date
A human-readable label like "~1850s" or "Day 3" — shown on cards and the calendar.
Datetime
ISO 8601 timestamp used for sorting and machine comparison.
On the timeline list, events with ranges show a band rather than a single dot. Overlapping bands between consecutive events are instant visual signals of contested or ambiguous sequencing.
Every timeline entity has a confidence score from 1 (very uncertain) to 5 (highly reliable). The meaning of each level is room-configurable— you define what "5" means in your context.
To customize labels: Go to Settings → Timeline Settings → Confidence Label Customization. Enter what each level means for your room type. These labels appear on timeline cards, detail pages, and in AI analysis prompts.
The same set of events can be arranged into multiple plausible orderings, each implying a different theory of what happened. Scenarios let you maintain competing timelines simultaneously.
Named scenarios
Each scenario has a name ("Theory A", "Prosecution Case", "Draft 2") and optional description. The active scenario is your current working theory.
Fork a scenario
Create a copy of an existing scenario to explore an alternative ordering without losing the original.
Compare scenarios
Side-by-side view showing which events are placed differently, which are excluded, and the biggest positional shifts. AI can identify the pivotal difference.
Archive scenarios
Archive theories that have been ruled out, with a note explaining why. Archived scenarios remain viewable but don't clutter the active workspace.
Every timeline event can be linked to a subject— a character, place, or object from your room's World entities. This creates parallel subject timelines that reveal convergence and separation points.
View by subject:On the timeline page, a pill strip shows all subjects referenced by events. Click a subject to filter the timeline to only that subject's events. Each subject gets a unique accent color.
Convergence points:When two subjects' time ranges overlap, it indicates they were potentially in the same place at the same time. In an investigation, this is alibi analysis. In fiction, it's character intersection.
Gaps: Periods where one subject has documented events but another has none — potentially suspicious or simply undocumented.
When you reorder events that have time ranges, the system automatically checks for logical contradictions. Conflicts appear as warning icons on affected event cards with explanatory tooltips.
IMPOSSIBLE
Adjacent events with non-overlapping ranges placed in an order that requires overlap. Example: Event A ends at 10pm, Event B starts at 8pm, but B is ordered after A.
TIGHT
Ranges overlap by less than 10% of the smaller range — technically possible but implausibly close.
ORDER GAP
Adjacent events both have ranges but zero time overlap — flagged for review but not necessarily wrong.
A summary strip appears at the top of the timeline when conflicts are detected: "2 ordering conflicts detected — review highlighted events."
When the same person, place, or thing appears across multiple sources under different names, you can group entities into an identity hypothesis.
Entity Groups
Soft-link entities that might be the same real-world entity ("John Smith" and "J. Smith" might be the same person). Groups appear as expandable cards on the bookshelf.
Typed Relationships
When entities are definitely different but connected: SUSPECT_OF, WITNESS_OF, SPOUSE_OF, COLLEAGUE_OF, and more.
AI Proximity Scanning
Click Scan for Connections on the bookshelf entities page. Sage AI analyzes all entities and proposes identity groups with rationale. Accept, reject, or edit each suggestion.
Merged Timeline
On any group's detail page, the Timeline tab shows all facts from all members on a shared chronological axis, color-coded by source entity.
Timeline events flagged as calendar events automatically appear on the room's calendar. This gives your team a visual overview of when events occurred — or might have occurred.
To mark an event as a calendar event: Check the Calendar Eventtoggle when creating or editing a timeline entity. The event's display date appears on the calendar.
Range events on the calendar: Events with range bands span multiple days on the calendar view, making uncertainty windows visually obvious.
Calendar status:Each event has a status ("historical", "scheduled", etc.) that affects its calendar appearance.
The room's Sage AI can analyze your timeline and provide insights that would take hours to compile manually.
Scenario Summary
Generate a plain-English summary of what must be true for a given scenario to be consistent — implied gaps, squeeze points, and dependencies. Uses your room's tag vocabulary and confidence labels.
Conflict Explanation
Ask Sage to explain any detected conflict in natural language and suggest what additional information would resolve the ambiguity.
Scenario Comparison
Give Sage two scenarios and it identifies the pivotal difference: which single event's placement is the key disagreement.
Narrative Synthesis
On any entity group's Analysis tab, click Refresh Analysis to generate a structured narrative: Summary, Evidence For, Evidence Against, and Unresolved Questions.
Chat with Timeline Context
When chatting with Sage, use @timelineto inject your active scenario's ordered event list into the conversation. Sage draws on your room's entire knowledge corpus.
Here's a recommended workflow for mystery investigation rooms. This pattern works for cold cases, journalism, historical research, and any room where reconstructing events from incomplete evidence is the core task.
Set up your room
Create an Investigation-type study room. Go to Settings → Timeline Settings and define your tag vocabulary (anchor, estimated, disputed, sentinel-start, sentinel-end, gap) and confidence labels (what 1–5 means in your context).
Build your world
Create a World with your known entities — characters (suspects, victims, witnesses), places (crime scenes, locations of interest), and timeline events (known facts with dates or date ranges). Import the World into your room in IMPORT mode.
Upload source documents
Upload police reports, witness statements, news articles, and other documents to the bookshelf. The extraction pipeline identifies additional entities. Each document becomes a discussion thread.
Tag and score everything
Tag events as anchor (confirmed), estimated (approximate), disputed (conflicting accounts), or gap (unexplained period). Set confidence scores. Events tagged as anchor with high confidence become the backbone of your timeline.
Create identity groups
When the same person appears under different names across sources, group those entities. Use Scan for Connections to let Sage propose groups automatically. Review, accept, or reject each suggestion.
Build competing scenarios
Create a baseline scenario with events in their default order. Fork it to create "Theory A" and "Theory B" with different event orderings. Use conflict detection to identify logical contradictions. Use scenario comparison to understand what differs.
Layer by subject
Assign subjects to timeline events. Use the View by subject filter to isolate individual timelines. Look for convergence points — when were two subjects in the same place at the same time? Look for gaps — who has an unaccounted-for period?
Analyze with Sage
Use scenario summaries to validate each theory. Ask Sage to explain detected conflicts. Chat with @timeline context to draw on your room's entire knowledge base. Use group narrative synthesis to understand identity hypotheses.
Archive and converge
As new evidence arrives, archive scenarios that have been ruled out (with notes explaining why). Eliminate impossible orderings. The surviving scenario represents your best current understanding — until new evidence suggests otherwise.