Cogitan
Documentation menu

Quickstart

Getting started

Read what may be believed before computing anything: the capability manifest states, for every function, its ground truth, its measured error against that ground truth, where it is trusted, and when it refuses.
Each step is a CLI command and the same request with curl. The curl lines read $FLUXUS_API and $COGITAN_API_KEY; set them as in Getting started. The CLI runs the engine in-process and needs neither.

  1. Set the base URL and key

    Every curl on these pages uses two variables. FLUXUS_API is the hosted API, and COGITAN_API_KEY is a cog_sk_ key from an approved account. Against a local engine, set FLUXUS_API to http://127.0.0.1:8787/api instead; it needs no key.

  2. Read the capability manifest

    Free on the hosted API. The CLI prints the same manifest from the engine in-process.

    CLI
    fluxus capabilities
    curl
    curl -s "$FLUXUS_API/capabilities" -H "Authorization: Bearer $COGITAN_API_KEY"
    GET/v1/fluxus/capabilitiesfluxus capabilities reference

  3. See which engines this host can serve

    Each engine with its version and validation card.

    No CLI command; `fluxus qubit status` covers the qubit engines.

    curl
    curl -s "$FLUXUS_API/engines" -H "Authorization: Bearer $COGITAN_API_KEY"
    GET/v1/fluxus/engines

  4. Ask one question

    A TransmonCross qubit with 90% intervals against HFSS.

    CLI
    fluxus qubit cross --cross-length-um 240 --claw-length-um 215 --ground-spacing-um 10 --junction-lj-nh 9.686 --length-um 3900 --coupling-length-um 225 --intervals
    curl
    curl -s "$FLUXUS_API/qem/cross" -H "Authorization: Bearer $COGITAN_API_KEY" \
  -H 'content-type: application/json' \
  -d '{"cross_length_um": 240, "claw_length_um": 215, "ground_spacing_um": 10, "junction_lj_nh": 9.686, "length_um": 3900, "coupling_length_um": 225, "intervals": true}'
    POST/v1/fluxus/qem/crossfluxus qubit cross reference

Generated from Fluxus 0.3.0 at commit b9aa57d on 2026-10-20 by python -m fluxus.docs_bundle.