URD ATLAS
Account
ChainsStatusTrack RecordThresholdsGlossaryFAQAPI DocsMethodology
Account
Developer onboarding

Common research workflows
from sample artifacts to analysis

One concrete path from sample artifacts to useful analysis. Use this page together with the sample pack and schema reference to get to a first useful notebook quickly.
← Back to API Docs

Workflow 1 — regime-conditioned panel

Load Meta for all chains, extract status.label, confidence.confidence_score, and top drivers, and build a chain-state panel.

import json, pathlib
from pprint import pprint

base = pathlib.Path("sample-pack")
rows = []
for rel in [
    "ethereum/2026-03-31/meta.json",
    "arbitrum/2026-03-25/meta.json",
    "ethereum/2025-04-21/meta.json",
]:
    row = json.loads((base / rel).read_text())
    rows.append({
        "chain": row["chain"],
        "date": row["date"],
        "label": row["status"]["label"],
        "confidence": row["confidence"]["confidence_score"],
        "top_driver": row["regime"]["drivers"][0]["metric"] if row["regime"].get("drivers") else None,
    })

pprint(rows)

Workflow 2 — verify a determinism hash

Use the example on Verification & Evidence Pack to prove that the named regime payload is internally consistent.

Workflow 3 — parse confidence-aware state changes

Treat UNKNOWN/DEGRADED as a first-class state, not as a missing row, and keep freshness and confidence separate from label interpretation.

Downloads