bridgic-amphibious

Bridgic Amphibious

Dual-mode agent framework: agents operate in LLM-driven (on_agent) and deterministic (on_workflow) modes with automatic fallback between them.

Dependencies

A bridgic-amphibious project requires the following packages:

Package	Description
bridgic-core	Core framework (Worker, Automa, GraphAutoma)
bridgic-amphibious	Dual-mode agent framework
bridgic-llms-openai	LLM provider (only required for AGENT / AMPHIFLOW modes)
python-dotenv	.env file loading

Before using this package, you need to install the dependencies by using the provided install script:

bash "skills/bridgic-amphibious/scripts/install-deps.sh" "$PWD"

The script checks uv availability, initializes a uv project if needed, installs any missing packages via uv add, and runs uv sync to finalize the environment. When it exits successfully the project is fully initialized and ready to use — no manual uv add / uv sync follow-up is required.

LLM Setup

Amphibious agents accept a BaseLlm instance with astructure_output protocol from a bridgic LLM provider package. The LLM is required for AGENT and AMPHIFLOW modes; pure WORKFLOW mode can run without one.

from bridgic.llms.openai import OpenAILlm, OpenAIConfiguration

llm = OpenAILlm(
    api_key="your-api-key",
    api_base="https://api.openai.com/v1",  # or custom endpoint
    configuration=OpenAIConfiguration(model="gpt-4o", temperature=0.0),
)

Other providers with same protocol: bridgic.llms.vllm.VllmServerLlm (self-hosted vLLM).

Quick Start

from bridgic.amphibious import (
    AmphibiousAutoma, CognitiveContext, CognitiveWorker, think_unit,
)
from bridgic.core.agentic.tool_specs import FunctionToolSpec

async def get_weather(city: str) -> str:
    """Get weather for a city."""
    return f"Sunny, 22°C in {city}"

class WeatherAgent(AmphibiousAutoma[CognitiveContext]):
    planner = think_unit(
        CognitiveWorker.inline("Look up weather and provide a summary."),
        max_attempts=5,
    )
    async def on_agent(self, ctx: CognitiveContext):
        await self.planner

agent = WeatherAgent(llm=llm, verbose=True)
result = await agent.arun(
    goal="Check the weather in Tokyo and London.",
    tools=[FunctionToolSpec.from_raw(get_weather)],
)