1
Shape the pilot
Settle the selected play situations, interfaces, and a finite way to evaluate the pilot.
Planning openSettle pilot mechanics/interfaces and the finite evaluation envelope.
Public roadmap
Building a dependable engine for persistent, human-led adventure.
Veylforge is creating the connective engine behind story-first play: an AI GM can interpret what people propose, while reviewed rules validate supported actions, save consequences, and return results the GM can use in the story.
The path ahead
These phases show dependency order rather than calendar promises. Work moves forward when the needed evidence and decisions are ready; dates are intentionally not invented before that planning is settled.
2
Make shared rules dependable
Turn reviewed rule definitions into repeatable, saved results through the existing engine.
Active; Block 8 needs repairRepair B8-R1/R2 and verify closure, then finish remaining shared execution/reuse; keep WI14 open.
3
Bring the rule path into connected play
Use the proven path with the selected gameplay foundation once the shared work is ready.
On holdEstimate after selected mechanics and shared-path readiness.
4
Prepare the real play connection
Join setup, context, recovery, and the operational path needed for a responsible trial.
OpenEstimate after the pilot/evaluation envelope and domain interfaces are selected.
5
Run a representative pilot
Observe selected play through the actual AI GM and Discord, with human choices and consequences preserved.
OpenEstimate after setup and operational arrangements.
6
Review the whole result
Bring together the evidence, independent review, and any repairs before deciding what is ready next.
OpenAssemble whole-effort evidence, independent review, and any identified repair work.