Data Explorer

Snapshot date

2026-11-05

Date range

2026-08-06 to 2026-11-04

Source table

bigquery-public-data.crypto_near_mainnet_us.transactions

Rows

91

Estimated bytes processed

1.84 GB

Data file

data/daily_metrics.csv

SQL (sql/daily_metrics.sql)

-- NEARPulse AI: daily transactions and distinct transaction signers
-- Source: bigquery-public-data.crypto_near_mainnet_us.transactions
-- Date range: 2026-08-06 to 2026-11-04 (inclusive)
-- Note: block_date is the partition column, so the date filter limits cost.

SELECT
  block_date,
  COUNT(*) AS transaction_count,
  COUNT(DISTINCT signer_account_id) AS distinct_signers
FROM `bigquery-public-data.crypto_near_mainnet_us.transactions`
WHERE block_date BETWEEN DATE '2026-08-06' AND DATE '2026-11-04'
GROUP BY block_date
ORDER BY block_date

Metric definitions

Transactions: Number of rows in the transactions table for that block_date.

Distinct signers: Number of distinct signer_account_id values among that day's transactions. This is NOT all active accounts and NOT unique users.

Known observations (unverified causes)

• Distinct signers were elevated from 2026-10-02 to 2026-10-07: 118077, 176719 (10-03), 177196 (10-04), 175774 (10-07) and 124809 (10-06) against a snapshot median of 70014. Cause not verified.
• Highest daily transaction counts: 2026-09-27 (1217209), 2026-10-01 (1151492), 2026-09-28 (1129793) against a snapshot median of 691420. Cause not verified.
• 2026-11-03 and 2026-11-04 show a sharp rise in distinct signers versus the days before (48060 on 11-02, 72343 on 11-03, 79164 on 11-04). Cause not verified.

Snapshot date

2026-11-06

Date range

2026-08-06 to 2026-11-04

Source table

bigquery-public-data.crypto_near_mainnet_us.receipt_actions

Rows

91

Estimated bytes processed

not recorded

Data file

data/function_calls_daily.csv

SQL (sql/function_calls_daily.sql)

-- NEARPulse AI: daily function-call actions, distinct receiver contracts, top-5 concentration
-- Source: bigquery-public-data.crypto_near_mainnet_us.receipt_actions
-- Date range: 2026-08-06 to 2026-11-04 (inclusive)
-- Note: one row per action, so one transaction can produce several function-call actions.
-- "distinct_contracts" = receiver accounts with at least one FUNCTION_CALL action that day.

WITH per_contract AS (
  SELECT
    block_date,
    receipt_receiver_account_id AS contract_account,
    COUNT(*) AS calls
  FROM `bigquery-public-data.crypto_near_mainnet_us.receipt_actions`
  WHERE block_date BETWEEN DATE '2026-08-06' AND DATE '2026-11-04'
    AND action_kind = 'FUNCTION_CALL'
  GROUP BY block_date, contract_account
),
ranked AS (
  SELECT
    *,
    ROW_NUMBER() OVER (PARTITION BY block_date ORDER BY calls DESC) AS rn
  FROM per_contract
)
SELECT
  block_date,
  SUM(calls) AS function_call_actions,
  COUNT(*) AS distinct_contracts,
  SUM(IF(rn <= 5, calls, 0)) AS top5_calls
FROM ranked
GROUP BY block_date
ORDER BY block_date

Metric definitions

Function-call actions: Number of rows in receipt_actions with action_kind = FUNCTION_CALL for that block_date. One transaction can produce several actions, so this is NOT a transaction count.

Distinct receiver contracts: Number of distinct receipt_receiver_account_id values with at least one FUNCTION_CALL action that day. This is NOT a complete measure of active contracts.

Top-5 calls: Function-call actions received by the 5 most-called receiver accounts that day.

Known observations (unverified causes)

• 2026-10-10, 2026-10-11 and 2026-10-15 had 7455968, 12700267 and 11615182 function-call actions, against a snapshot median of 1482979. The top 5 receiver accounts held 69.5%, 87.4% and 91.4% of actions on those days (snapshot median share: 47.2%). Distinct contracts on those days (1477, 1524, 1476) did not rise. Cause not verified.
• Distinct receiver contracts were between 832 and 921 per day from 2026-08-06 to 2026-08-13, and between 1137 and 1523 per day from 2026-10-16 to 2026-10-31. The reason for the rise is not verified.

Source table

bigquery-public-data.crypto_near_mainnet_us.receipt_actions

Days

2026-10-10, 2026-10-11, 2026-10-15

Rows

15

Data file

data/spike_days_top5.json

SQL (sql/spike_days_top5.sql)

-- NEARPulse AI: top 5 receiver accounts by function-call actions on three spike days
-- Source: bigquery-public-data.crypto_near_mainnet_us.receipt_actions
-- Days: 2026-10-10, 2026-10-11, 2026-10-15
-- share_pct = account's calls / all function-call actions that day

WITH per_contract AS (
  SELECT
    block_date,
    receipt_receiver_account_id AS contract_account,
    COUNT(*) AS calls
  FROM `bigquery-public-data.crypto_near_mainnet_us.receipt_actions`
  WHERE block_date IN (DATE '2026-10-10', DATE '2026-10-11', DATE '2026-10-15')
    AND action_kind = 'FUNCTION_CALL'
  GROUP BY block_date, contract_account
),
ranked AS (
  SELECT
    *,
    ROW_NUMBER() OVER (PARTITION BY block_date ORDER BY calls DESC) AS rn,
    SUM(calls) OVER (PARTITION BY block_date) AS day_total
  FROM per_contract
)
SELECT
  block_date,
  rn AS rank,
  contract_account,
  calls,
  day_total,
  ROUND(100 * calls / day_total, 1) AS share_pct
FROM ranked
WHERE rn <= 5
ORDER BY block_date, rn

Result: five most-called receiver accounts per day

What these accounts are, and why they received so many calls, was not investigated.

How to reproduce

1. Open console.cloud.google.com/bigquery (the free sandbox is enough).
2. Click Copy query on one of the queries above and paste it into a new query tab.
3. Check the "This query will process" estimate, then click Run.
4. Compare the result with the matching file in the data folder of the repository.