Pipeline & ask()

POST/api/v1/pipelines/run

ask() and run_pipeline() POST thresholds and options. Module on/off comes from governance policies (effective_modules), not request booleans.

out = ts.ask(
    "Summarize in 3 bullets.",
    provider="openai",
    model="gpt-4o",
    cache_similarity_threshold=0.85,
)
print(out.text, out.metrics.cost_usd)

Enable cache/RAG/PII first via governance policies. For field-level parity, see the full method reference (Pipeline options card).