CRB is a hybrid actor/FSM/task framework for Rust. Create autonomous agents, workflows, and AI pipelines with minimal boilerplate.
View on GitHub →
Composable blocks for autonomous systems
Seamlessly combine actors, finite state machines, sync/async tasks, and supervisors. Choose the right abstraction for each component.
Build once, run everywhere. Full no-std support means your agents work in browsers, embedded systems, and servers without modification.
Extensible architecture through traits. Implement your own actors, FSMs, and tasks while leveraging the framework's infrastructure.
Decompose complex workflows into graph structures with nodes as operations and edges as state transitions. Build cyclical flows, conditional branching, and multi-agent orchestration patterns.
Build large-scale systems from small, focused components. Each piece is independently testable and replaceable. Swap implementations without changing the overall structure.
Powers a complete educational platform in production: interactive book reader app, AI coding agent, content processing pipeline, and backend services.
Create actors, FSMs, or tasks using simple trait implementations. Define your agent's behavior and state transitions.
Connect components into workflows and pipelines. CRB handles message routing, lifecycle, and coordination.
Run your agents in any environment: web servers, WASM apps, CLI tools, or embedded systems.
Agents utilize all available system resources. Spawn multiple instances when load increases. Scale horizontally across machines or vertically within a single process.
CRB is source available under BSL license and available on GitHub
Get Started →We're constantly evolving CRB based on community feedback
Run agents across multiple machines with transparent message routing
Drag-and-drop interface for designing agent workflows
Update agent behavior without restarting the system
Shape the architecture — don't just adopt the framework
We collaborate with a small group of engineering organizations to evolve CRB based on real production use and long-term system requirements.
Your technical requirements and constraints inform feature priorities, API design, and core architectural decisions.
Review upcoming changes, breaking updates, and new capabilities before they become generally available.
Direct communication with core maintainers to discuss production challenges, performance bottlenecks, and integration patterns.
CRB evolves through controlled, incremental change — guided by real organizational needs, not hypothetical use cases.
Participation is intentionally structured to preserve architectural clarity, signal quality, and meaningful technical dialogue.
Limited to 20 organizations.
CRB is source available under BSL license and available on GitHub
Get Started →
