aevatar-agent-framework
A distributed agent development framework based on the Actor model, enabling a single codebase to adapt to three runtime environments: local, high-performance, and distributed.
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: aevatar-agent-framework. It is used for: A distributed agent development framework based on the Actor model, enabling a single codebase to adapt to three runtime environments: local, high-performance, and distributed. Full Skill content: https://321skill.com/skills/aevatar-agent-framework/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
The Aevatar Agent Framework aims to address the pain point in agent development where code is deeply coupled with the runtime environment. Developers often need to write different code for local debugging, high-performance scenarios, or distributed deployment, leading to high development and maintenance costs.
Through abstraction via the Actor model, this framework allows developers to define agents using the same C# codebase. By simply switching one line of configuration in the dependency injection container (e.g., selecting Local, ProtoActor, or Orleans runtime), agents can seamlessly run in different environments—from local processes to large-scale distributed clusters—without modifying business logic.
It is well-suited for backend, full-stack, and agent developers who need to build scalable, high-performance agent systems. Whether developing AI applications requiring rapid local prototyping or constructing distributed agent services that must handle high concurrency and fault tolerance, this framework enables efficient development.
A crucial constraint to note before use: all state and event types requiring serialization must be defined using Protocol Buffers (.proto files). The framework enforces type safety and cross-platform compatibility; using plain C# classes directly will cause runtime failures. It is recommended to start with the official examples to familiarize yourself with the Protobuf integration process.
Key Features
The core differentiator lies in its "one codebase, multiple runtimes" architectural design. Similar frameworks typically bind to a single runtime (e.g., supporting only local Actors or only Orleans). In contrast, Aevatar, through an abstraction layer, enables the same agent implementation to seamlessly switch between local development, high-performance ProtoActor clusters, and distributed Orleans environments, significantly enhancing code portability and deployment flexibility.
Limitations
Not suitable for projects unfamiliar with Protocol Buffers or those wishing to freely define serialization data structures, as the framework mandates that all serializable types must be defined using Protobuf.
FAQ
Why is Protocol Buffers mandatory? Why can't plain C# classes be used?
This is a strict framework constraint designed to ensure serialization consistency, type safety, and performance across the three runtimes (Local/ProtoActor/Orleans). Orleans transmission uses byte arrays, ProtoActor requires cross-language support, and the local runtime requires deep copying. Only Protobuf can satisfy all these scenarios simultaneously.
How should I choose among the three runtimes?
Use LocalRuntime for local development and debugging; choose ProtoActorRuntime for extremely high single-machine performance; and select OrleansRuntime for building elastic, scalable distributed systems.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/aevatar-agent-framework/raw/index.md to read the original Skill definition (Markdown format) for aevatar-agent-framework, and install it according to the instructions.
Raw Markdown URL for AI: /skills/aevatar-agent-framework/raw/index.md