Stella Multi-Agent Framework

The Stella Multi-Agent Framework builds scalable agent-based workflows with your model of choice, including the ability to deploy Twitter agents and Warpcast agents. The framework is designed to perform robustly with IBM Granite and Llama 3.x models, and we're actively working on optimizing its performance with other popular LLMs.

Our goal is to empower developers to adopt the latest open-source and proprietary models with minimal changes to their current agent implementation.

Getting started

Installation

shell npm install Stella-agent-framework

or

shell yarn add Stella-agent-framework

Example

```ts import { StellaAgent } from "Stella-agent-framework/agents/Stella/agent"; import { OllamaChatLLM } from "Stella-agent-framework/adapters/ollama/chat"; import { TokenMemory } from "Stella-agent-framework/memory/tokenMemory"; import { DuckDuckGoSearchTool } from "Stella-agent-framework/tools/search/duckDuckGoSearch"; import { OpenMeteoTool } from "Stella-agent-framework/tools/weather/openMeteo";

const llm = new OllamaChatLLM(); // default is llama3.1 (8B), it is recommended to use 70B model

const agent = new StellaAgent({ llm, // for more explore 'Stella-agent-framework/adapters' memory: new TokenMemory({ llm }), // for more explore 'Stella-agent-framework/memory' tools: [new DuckDuckGoSearchTool(), new OpenMeteoTool()], // for more explore 'Stella-agent-framework/tools' });

const response = await agent .run({ prompt: "What's the current weather in Las Vegas?" }) .observe((emitter) => { emitter.on("update", async ({ data, update, meta }) => { console.log(Agent (${update.key}) 🤖 :, update.value); }); });

console.log(Agent 🤖 :, response.result.text); ```

Local Installation

[!NOTE]

yarn should be installed via Corepack (tutorial)

1. Clone the repository git clone git@github.com:i-am-Stella/Stella-agent-framework.
2. Install dependencies yarn install --immutable && yarn prepare.
3. Create .env (from .env.template) and fill in missing values (if any).
4. Start the agent yarn run start:Stella (it runs /examples/agents/Stella.ts file).

➡️ All examples can be found in the examples directory.

➡️ To run an arbitrary example, use the following command yarn start examples/agents/Stella.ts (just pass the appropriate path to the desired example).

📦 Modules

The source directory (src) provides numerous modules that one can use.

| Name | Description | | ------------------------------------------------ | ------------------------------------------------------------------------------------------- | | agents | Base classes defining the common interface for agent. | | llms | Base classes defining the common interface for text inference (standard or chat). | | template | Prompt Templating system based on Mustache with various improvements. | | memory | Various types of memories to use with agent. | | tools | Tools that an agent can use. | | cache | Preset of different caching approaches that can be used together with tools. | | errors | Error classes and helpers to catch errors fast. | | adapters | Concrete implementations of given modules for different environments. | | logger | Core component for logging all actions within the framework. | | serializer | Core component for the ability to serialize/deserialize modules into the serialized format. | | version | Constants representing the framework (e.g., latest version) | | emitter | Bringing visibility to the system by emitting events. | | internals | Modules used by other modules within the framework. |

To see more in-depth explanation see overview.

Key Features

• 🤖 AI agents: Use our powerful Stella agent refined for Llama 3.1 and Granite 3.0, or build your own.
• 🛠️ Tools: Use our built-in tools or create your own in Javascript/Python.
• 👩‍💻 Code interpreter: Run code safely in a sandbox container.
• 💾 Memory: Multiple strategies to optimize token spend.
• ⏸️ Serialization Handle complex agentic workflows and easily pause/resume them without losing state.
• 🔍 Instrumentation: Use Instrumentation based on Emitter to have full visibility of your agent’s inner workings.
• 🎛️ Production-level control with caching and error handling.
• 🔁 API: Integrate your agents using an OpenAI-compatible Assistants API and Python SDK.
• 🖥️ Chat UI: Serve your agent to users in a delightful UI with built-in transparency, explainability, and user controls.