01
Start free.
Open the Analyst Kit and use a public regime-calendar CSV endpoint directly. You do not need an account or a manual download for the first test.
Open Analyst Kit →Getting started
Test one file, join it to your data, then subscribe only when the workflow works.
This page assumes no prior Urd Atlas setup. Follow it from top to bottom to run your first CSV join or copy your first readable context note.
No price data · No forecasts · No recommendations
02
Join on date + chain.
Your table needs a date column and a chain column. Urd Atlas adds the published state for that same daily row.
03
Subscribe for delivery.
After the join is useful, choose one chain for $49/mo or all four chains for $149/mo to receive authenticated daily files.
View pricing →The four files
Pick the layer that matches your workflow.
The examples below use the raw published JSON shapes. The Analyst Kit regime-calendar endpoint is a separate flattened CSV contract designed for direct date-and-chain joins.
Meta
Use the regime and confidence score as the daily state row you join to your own table.
meta["status"]["label"], meta["confidence"]["confidence_score"]Gold
Use the measured chain activity behind the label when you want to inspect the evidence yourself.
gold["date"], gold["chain"], gold["tx_count_daily"]Derived
Use pre-computed rolling windows when you do not want to rebuild feature engineering.
derived["derived"]["metrics"]["tx_count_daily__ma7"]Briefs
Use the plain-language sentence when a report or dashboard needs readable context.
brief["one_liner"]Join example
Attach Urd Atlas to your own daily table.
The example reads Ethereum's public regime calendar directly from Urd Atlas, so only your own daily table needs to exist locally. Change the chain in the URL when you want to test BTC, ARB or BASE.
Urd Atlas uses 0.40 as the publication confidence gate. The 0.70 filter below is intentionally stricter and is only an example of a downstream analyst choosing to work with higher-confidence rows.
For historical analysis, `observation_date` tells you which day the network state describes; it does not prove that the row was available on that date. Do not treat the current historical calendar as point-in-time data for decision simulation. Use an availability timestamp once point-in-time/vintage exports are enabled.
import pandas as pd
urd = pd.read_csv(
"https://urdatlas.com/api/v1/analyst-kit/ethereum/regime-calendar"
)
my_data = pd.read_csv("my_daily_metrics.csv")
joined = my_data.merge(
urd[["observation_date", "chain", "regime", "confidence_score"]],
left_on=["date", "chain"],
right_on=["observation_date", "chain"],
how="left",
)
# Optional downstream filter; stricter than the Urd Atlas publication gate of 0.40.
usable = joined[joined["confidence_score"] >= 0.70]No-code path
Use Explorer and Briefs when you do not have a pipeline.
Open Explorer to read the current published state. Use Briefs when you need a short sentence for a report, internal note or dashboard.