DOCS / PUBLIC GUIDE
Working concepts.
A practical overview of the ideas shaping Vivacity. This is not final API documentation.
Private technical access
REVISIONSeptember 2026
01 / OVERVIEW
Vivacity is exploring infrastructure for agents that operate in changing environments.
Most agent software is built around requests and responses. Real activity is different. Conditions change. Actions have side effects. Information arrives late or only in part. Vivacity is focused on the systems required to keep an agent grounded through that process.
The public work is early and available through selected technical partnerships. Concepts and interfaces may change as the research develops.
02 / WORLD MODELS
A world model predicts. An agent still needs a system for acting on that prediction.
Vivacity sits in the unresolved space between learned dynamics and operational software. The ambition is not another general-purpose world model. It is a runtime boundary where model rollouts, exact solvers, physics engines, and domain systems can inform the same agent decision without erasing their differences.
World state plus action is routed through an execution process to produce a candidate next state. This notation is conceptual.
03 / ENVIRONMENTS
An environment is a state space with memory.
It includes the conditions an agent can observe, the constraints it must respect, and the variables that continue to evolve while the agent is working. A useful representation preserves identity, time, lineage, and admissible transitions.
- ContinuityState survives beyond one inference turn.
- Constraint surfaceThe environment defines the boundary of admissible change.
- LineageEvery candidate state retains its causal origin.
04 / ACTIONS
An action is a proposed state transition.
Vivacity treats action as something that can be typed, checked, executed, and evaluated. Preconditions define when a transition is admissible. Postconditions describe the evidence expected after execution.
05 / OBSERVATIONS
Observation is a projection. State is the source.
An observation operator maps a larger environment into the partial view an agent needs. That view may be visual, geometric, symbolic, or multimodal. It can be delayed or noisy without redefining the underlying state.
The persistent variables, constraints, relations, and time index.
A task-conditioned projection for the next decision.
06 / EXECUTION
Route each transition to the physics it requires.
Closed-form dynamics may call for an exact solver. Contact-rich mechanics may require a physics engine. Open environments may benefit from learned rollouts. Private systems may use domain logic. The runtime should preserve a common causal trace while keeping each executor's evidence intact.
No third-party integrations are claimed here. Execution paths are research directions and partnership topics.
07 / STATUS AND LIMITS
A technical direction, not a fictional release.
- No public SDK yetInterfaces remain subject to change.
- No published benchmarks yetMeasurements will be shared only when they are meaningful.
- No implied integrationsNamed third-party systems are not presented as connected products.