dynabots-core

Core primitives for the DynaBots orchestration framework family.

Overview

dynabots-core provides the foundational building blocks shared across all DynaBots orchestration frameworks:

• Protocols: Agent, LLMProvider, Judge, Tool - interfaces that define contracts
• Value Objects: TaskResult, LLMMessage, LLMResponse - immutable data structures
• Providers: Ready-to-use LLM adapters for OpenAI, Ollama, Anthropic

Installation

# Core only (bring your own LLM client)
pip install dynabots-core

# With OpenAI support
pip install dynabots-core[openai]

# With Ollama (local LLMs)
pip install dynabots-core[ollama]

# With all providers
pip install dynabots-core[all]

Quick Start

Define an Agent

from dynabots_core import Agent, TaskResult

class MyAgent:
    @property
    def name(self) -> str:
        return "MyAgent"

    @property
    def capabilities(self) -> list[str]:
        return ["search", "summarize"]

    async def process_task(self, task_description: str, context: dict) -> TaskResult:
        # Your agent logic here
        result = await self._do_work(task_description)
        return TaskResult.success(
            task_id=context["task_id"],
            data=result
        )

Use an LLM Provider

from dynabots_core.providers import OllamaProvider

# Local LLM via Ollama
llm = OllamaProvider(model="qwen2.5:72b")

response = await llm.complete([
    LLMMessage(role="system", content="You are a helpful assistant."),
    LLMMessage(role="user", content="Hello!")
])
print(response.content)

Create a Judge (for Orc!!)

from dynabots_core import Judge, Verdict

class LLMJudge:
    def __init__(self, llm: LLMProvider):
        self.llm = llm

    async def evaluate(self, task: str, submissions: list[dict]) -> Verdict:
        # Compare submissions and pick a winner
        prompt = f"Task: {task}\n\nSubmission A: {submissions[0]}\nSubmission B: {submissions[1]}\n\nWhich is better?"
        response = await self.llm.complete([LLMMessage(role="user", content=prompt)])
        return Verdict(winner=..., reasoning=response.content)

Framework Family

This package is part of the DynaBots framework family:

Protocols

Agent

class Agent(Protocol):
    @property
    def name(self) -> str: ...

    @property
    def capabilities(self) -> list[str]: ...

    async def process_task(self, task_description: str, context: dict) -> TaskResult: ...

LLMProvider

class LLMProvider(Protocol):
    async def complete(
        self,
        messages: list[LLMMessage],
        temperature: float = 0.1,
        max_tokens: int = 2000,
        json_mode: bool = False,
    ) -> LLMResponse: ...

Judge

class Judge(Protocol):
    async def evaluate(self, task: str, submissions: list[dict]) -> Verdict: ...

Tool

class Tool(Protocol):
    @property
    def name(self) -> str: ...

    @property
    def description(self) -> str: ...

    @property
    def parameters_schema(self) -> dict: ...

    async def execute(self, **kwargs) -> Any: ...

License

MIT