Plantangenet is a framework for creating bounded, policy-enforced worlds where sharing is negotiated, partial, auditable, and ephemeral as a core feature.
It gives developers, system architects, and world-builders the tools to model trust, and privacy as first-class system features.
If you'd like to dive into technical details right away, see the Technical README.
Modern systems often assume static permissions and all-or-nothing sharing.
Plantangenet flips that model:
- Negotiated Disclosure
Define exactly what you share, with whom, and at what level of detail. - Enforceable Policies
Machine-enforced rules for access, roles, and conditions. - Ephemeral Memory
Attention-shaped, fading storage that models forgetting as a feature. - Economic Incentives
Control retention and access with explicit costs (Dust).
- Hackers building strange, constrained, privacy-aware systems
- CTOs designing federated or zero-trust architectures
- Game and economy designers modeling negotiated trade and partial disclosure
- Collaborative creators who need nuanced sharing contracts
- Private federated services with role-based views
- Multiplayer games with secret negotiation
- Attention-priced data markets
- Privacy-conscious simulations
- Collaborative apps with time-bounded sharing
- Distributed ML pipelines with native tensor support and GPU-accelerated data transformation
- Boundaries are explicit.
- Trust is negotiated.
- Memory is shaped by attention.
- Privacy is enforced by policy.
Plantangenet includes:
- Sessions with policy-bound lifecycle
- Agents modeling roles and permissions
- Cursors for fine-grained data focus
- Compositors for partial views
- Membry for attention-shaped, fading storage
- GPU-accelerated "shaders" for distributed data transformation
- Native tensor support for ML workloads
- Seamless integration with popular Python data tools like NumPy and pandas
- A pluggable Policy interface
- Economic incentives via Dust
- Observables and Output: Compositors and comdecs (loggers, dashboards, streamers) subscribe to session or agent state and update outputs in real time. This enables live dashboards, streaming, and pluggable output formats with minimal coupling.
Plantangenet uses an observable pattern: compositors and comdecs subscribe to session or agent state, and automatically update outputs (logs, dashboards, streams) when state changes. This enables real-time dashboards, live MJPEG streaming, and pluggable output formats with minimal coupling.
- Agents expose state and rendering methods (e.g.
get_widget,__render__). - Compositors (like dashboards) collect and arrange widgets from agents.
- Observables (comdecs, loggers, streamers) subscribe to state and update outputs in real time.
- Session coordinates all updates, policy checks, and output notifications.
Active, evolving, and open to collaboration.
Plantangenet is an independent, self-funded project. If you believe in building software that models trust, negotiation, and privacy as first-class features, please consider supporting development:
- Emotional Support Tier ($5/mo):
For those who want to see Plantangenet grow and receive regular updates. Your encouragement really matters. - On-Call Support Tier ($10,000/mo):
For institutions or projects that want on-call help designing or deploying negotiated systems.
More support tiers are coming soon for individuals and teams seeking closer collaboration, early access to design discussions, or custom integrations. If you have ideas or needs, reach out!
Your support and participation help keep Plantangenet independent, evolving, and open.
Hi! I'm Scott Russell (Queuetue), a systems architect and AI specialist with over 25 years of experience designing, scaling, and leading technical teams.
I specialize in turning early-stage ideas into resilient, high-performing systems. My expertise lies at the intersection of AI/ML, distributed systems, and data-intensive applications. I have deep experience with simulation, agent-based modeling, GPU-accelerated computing, and building infrastructure for complex, decentralized workloads.
My work is informed by natural systems thinking: applying principles of feedback, resilience, and sustainable complexity to both software and organizations. I am passionate about building the next generation of privacy-aware, distributed systems. Yes, for interested parties, I do consulting jobs on the side.
© 1998–2025 Scott Russell
SPDX-License-Identifier: MIT