Penpot Working Loop For Event Modeling And FnHCI
This note captures the current NEXUS working hypothesis for how Penpot can be used productively without becoming the final source of truth.
The short version is:
- yes, Penpot can help us build screens
- yes, Penpot can help us lay out paths and click through them with prototypes
- yes, Penpot can likely serve as a useful visual canvas for Event Modeling diagrams
- no, Penpot alone should not become the lasting semantic source of truth
The deeper source of truth should still live in structured NEXUS, FORGE, and FnHCI surfaces.
Why This Matters
There is an obvious catch-22 here:
- we want an Event Modeling and UI tool of our own
- we do not have that tool yet
- Penpot is good enough to help us move now
- but if we let Penpot own the meaning, we risk getting trapped inside the bridge tool
So the answer is not "Penpot or our own tool."
The answer is:
- use Penpot as a working canvas and bridge surface now
- keep the deeper language and semantics durable outside Penpot
- gradually turn the repeated useful parts into deterministic
FnToolsand FORGE surfaces
Penpot's Useful Roles Right Now
Penpot can already serve several legitimate roles.
Screen Authoring Surface
Penpot can hold:
- screen boards
- component libraries
- tokens and styling
- layout experiments
- variants and visual states
This is the most obvious current use.
Prototype Surface
Penpot can also help us:
- wire board-to-board navigation
- click through a path screen by screen
- test whether the sequence feels right for a human
- compare the modeled path to the actual visual flow
This is especially useful for PATH work where a human needs to feel the flow, not just read about it.
Event Modeling Canvas
Penpot should also be considered a candidate visual canvas for Event Modeling diagrams themselves.
That may include:
PATHboardsCommandSliceandViewSlicelayouts- event chains
- supporting annotations
- multi-fidelity views of the same path
This does not mean Penpot is the final Event Modeling tool.
It means Penpot can hold visual working artifacts while NEXUS is still forming the dedicated tool.
FnHCI Pressure Surface
Penpot can pressure the FnHCI design directly by making us answer:
- which parts are semantic primitives
- which parts are layout
- which parts are visual treatment
- which parts are runtime-specific
- which parts are only design-time convenience
That makes Penpot useful for finding the right abstraction level.
The Important Boundary
The visual Penpot artifact should not be the only source of truth.
The current intended structure is:
- business/domain truth
commands, events, policies, views, and
PATHsemantics FnHCItruth reusable interaction primitives and view composition semantics- Penpot visual working model boards, prototypes, components, variants, tokens, annotations
- runtime implementations
FnUI, Blazor, HTML, Android, iOS, and similar targets
That means:
- Penpot helps express and test the model
- Penpot does not own the model's deepest meaning
- prototypes help validate paths
- prototypes do not replace the path definition itself
The Working Loop
The current dogfooding loop should probably be:
- define or refine
PATH,CommandSlice,Event,ViewSlice, andViewin durable repo memory - represent the relevant screens and flow in Penpot
- wire Penpot prototype links so the path can be clicked through
- compare the Penpot path to the semantic path and reconcile mismatches
- extract or define the reusable
FnHCIprimitives implied by the Penpot components - implement the
FnUIruntime projection over those primitives - compare the resulting runtime output back to the Penpot design and the semantic path
This turns Penpot into a working collaborator surface instead of a hidden design island.
What Penpot Should Help Us Learn
If we use it well, Penpot should help us discover:
- what the real stable screen states are
- what the real
PATHtransitions are - which components are truly reusable
- which component variations are semantic versus merely stylistic
- which design ideas should become
FnHCIprimitives - where the Event Modeling tool eventually needs better purpose-built behavior than Penpot can provide
Suggested Penpot Areas
The current practical Penpot organization likely wants pages such as:
ComponentsScreensPATHSEventModels
And likely naming conventions such as:
Screen.NewSessionScreen.EntryFormPath1.1-NewSessionPath1.3-EntryFormButton.OptionButton.Counter
The exact conventions can evolve, but the principle should remain:
- names should help us map visual artifacts back to durable modeled concepts
How This Connects To FnHCI
The goal is not to derive HTML or Blazor directly from arbitrary Penpot shapes.
The better direction is:
- Penpot components pressure the primitive design
FnHCIowns the primitive modelFnUIand other runtimes render those primitives- Penpot remains a design/projection surface over the same deeper concepts
So if Penpot contains a Button.Counter component, that should help us discover:
- a likely
Buttonprimitive - a likely counter-style or increment/decrement affordance
- a likely set of state and content concerns
without making the Penpot component itself the deepest truth.
What Should Become Deterministic Later
Over time, this should become more toolable and less conversational.
Likely future deterministic surfaces include:
- Penpot board and component inspection helpers
- Penpot prototype extraction helpers
- Penpot-to-
PATHor Penpot-to-FnHCImapping helpers FnHCIprimitive catalogsFnUIprojection and comparison workflows- eventually a dedicated Event Modeling tool that carries these concepts more natively than Penpot
That is exactly the kind of progression FORGE is meant to encourage.
Current Rule Of Thumb
Use Penpot for:
- visualizing
- prototyping
- componentizing
- discussing
- pressuring the abstractions
Do not rely on Penpot alone for:
- semantic source of truth
- final path meaning
- business event definition
- long-term deterministic compiler behavior