☰ Documentation menu
Introduction
What is Provenova? Getting started
Concepts & workflows
Core concepts Capturing runs Reproduce & drift Result cards & badges Compliance & attestations Corpus & leaderboard
Using the app
Product tour (every page & button) Pricing FAQ
Libraries & reference
Libraries & downloads CLI reference API reference Frameworks reference Open schemas (qlprov) Deployment & self-hosting Licensing
Concepts & workflows

Result cards & badges

A Result Card is a public, citable page for a run — the shareable face of an otherwise internal record. Badges are the embeddable shields that advertise a result's maturity.

Publishing a card

From a run at /app/records/<run_id>, use Publish to create a Result Card. A card has:

  • a slug and public URL (/cards/<slug>),
  • a title and summary,
  • the run's provenance (backend, calibration, run_hash),
  • the result distribution,
  • visibility — private or public,
  • a free persistent identifier (PID), an optional real DOI minted on demand via Zenodo, and a license.

Cards can be unpublished (retracted) later.

Citing a result

Because a card carries a persistent identifier and an immutable run_hash, it can be cited unambiguously. The card page and API offer ready-made citations:

  • GET /api/v1/cards/<slug>/citation?format=<bibtex|csl|ris> — BibTeX (default), CSL JSON, or RIS.
  • GET /api/v1/cards/<slug> — machine-readable metadata (JSON).
  • GET /api/v1/cards/<slug>/embed — JSON of copy-paste snippets (a Markdown/HTML badge and an iframe tag).
  • GET /cards/<slug>/embed.html — the self-contained, iframe-embeddable Result Card widget itself (served with frame-ancestors *), also discoverable via oEmbed at GET /api/v1/oembed.

A card's free PID already satisfies the FAIR persistent-identifier control (FAIR-F1); an optional DOI — minted for free on demand via Zenodo, which also archives the provenance record — strengthens it with an external, globally-resolvable identifier. See Compliance.

Badges

Badges are shields.io-style images you can drop into a README, paper, or dashboard:

  • GET /badge/<slug>/<type>.svg — the badge as SVG.
  • GET /badge/<slug>/<type>.json — the same data as JSON (for automation).

The badge ladder

Badges reflect a result's maturity as a ladder — each rung is a stronger claim than the last:

  1. Recorded — the run is in the ledger with full provenance.
  2. Reproduced — it has been re-run and scored.
  3. Benchmarked — scored against the noiseless ideal and ranked within your workspace via Benchmark vs fleet.
  4. Compliant — it satisfies an enabled compliance framework.
  5. Audit-ready — backed by a signed attestation.

Rungs 1–3 are reachable on the Free plan for your own runs — capture, reproduce, then hit Benchmark vs fleet on the record page to record a benchmark entry and light the Benchmarked badge, no upgrade required. Rung 4 needs an enabled framework and rung 5 needs a signed attestation (included from the free Academic tier and all paid tiers — see the pricing FAQ).

External parties can even submit their own reproductions of a public card (POST /api/v1/cards/<slug>/reproductions) — independent verification that upgrades the card's Reproduced badge.

Next: Compliance & attestations.