GigaFlow CLI¶

Command-line client for GigaFlow — connect your LLM/agent traces to a GigaFlow backend and run Flow analysis on them: atomize each step, attribute information flow between atoms, score groundedness / relevance / fulfilment, and diagnose failures.

The CLI is a thin, zero-dependency client (Python standard library only). Your traces live in an observability platform (Arize Phoenix, Logfire, Braintrust, MLflow, W&B Weave) or are sent via OTLP; the backend does the compute; this CLI drives ingest → compute → inspection.

Install¶

pip install gigaflow

Configure¶

export GIGAFLOW_API_KEY=<your GigaFlow API key>
# Backend defaults to the hosted service (https://api.gigaflow.io/api/v1).
# For local dev: export GIGAFLOW_BACKEND_URL=http://localhost:8000/api/v1

gigaflow login (browser sign-in) or gigaflow setup persist these to ~/.gigaflow/config.json, so the exports are optional on later runs.

End-to-end in five commands¶

gigaflow setup                                   # pick your tracing tool, connect it, sync
gigaflow compute "SELECT trace_id FROM trace_metrics WHERE run_id IS NULL"
gigaflow inspect <trace_id>                      # open the browser Flow viewer
gigaflow query "SELECT trace_id, groundedness, total_cost_usd FROM trace_metrics ORDER BY total_cost_usd DESC LIMIT 20"

Where to next¶

Connect a trace source — Arize Phoenix, Logfire, Braintrust, MLflow, W&B Weave, or direct OTLP.
Querying — explore the trace_metrics view with SQL.
Transform configs — map raw vendor spans to GigaFlow primitives.
Changelog — release history.

The full command reference lives in the project README.