Skip to content

ontocast.integrations

Adapters that expose OntoCast to other agent frameworks.

Kept out of ontocast/tool/, which holds the stateful tools the ToolBox owns and injects. These are outward-facing wrappers around those.

Nothing here is imported by the pipeline itself, so langchain_core.tools stays off the cold import path of import ontocast.

Modules:

Name Description
langchain

Expose OntoCast capabilities as LangChain tools.

langgraph

Embed the OntoCast pipeline as a node in someone else's LangGraph.

schemas

Argument schemas for the LangChain tool wrappers.

serialize

Render OntoCast return values as text an LLM can read.

Attributes

__all__ = ['make_ontocast_node', 'ontocast_tool_diagnostics', 'ontocast_tool_names', 'ontocast_tools', 'text_in_turtle_out'] module-attribute

Functions:

__dir__()

Source code in ontocast/integrations/__init__.py
def __dir__() -> list[str]:
    return sorted([*globals(), *__all__])

__getattr__(name)

Resolve the integration entry points on first access.

Source code in ontocast/integrations/__init__.py
def __getattr__(name: str) -> Any:
    """Resolve the integration entry points on first access."""
    target = _LAZY_EXPORTS.get(name)
    if target is None:
        raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
    import importlib

    module_name, attribute = target
    value = getattr(importlib.import_module(module_name), attribute)
    globals()[name] = value
    return value