Skip to content

ontocast.prompt.web_grounding

Optional web-grounding surface for LLM prompts and schema generation.

persist_search_request(state, node, request, enabled)

Store LLM search request when enabled; reset to default when disabled.

Source code in ontocast/prompt/web_grounding.py 
66
67
68
69
70
71
72
73
74
75
76
def persist_search_request(
    state: UnitFactsState | UnitOntologyState,
    node: WorkflowNode,
    request: ExternalEvidenceRequest | None,
    enabled: bool,
) -> None:
    """Store LLM search request when enabled; reset to default when disabled."""
    if enabled and request is not None:
        state.set_external_evidence_request(node, request)
    else:
        state.set_external_evidence_request(node, ExternalEvidenceRequest())

search_guidelines_for(node, enabled)

Return node-specific search guideline prose, or empty when disabled.

Source code in ontocast/prompt/web_grounding.py 
59
60
61
62
63
def search_guidelines_for(node: WorkflowNode, enabled: bool) -> str:
    """Return node-specific search guideline prose, or empty when disabled."""
    if not enabled:
        return ""
    return _GUIDELINES_BY_NODE.get(node, "")