Skip to content

ABOUT / METHOD

NO MAGIC.
JUST EXECUTION.

VgeoCPU does not predict markets.

It reads publicly available execution data and converts it into a compact compute profile. Every displayed metric is traceable back to source transactions or blocks.

If a value cannot be calculated from fetched data it prints UNAVAILABLE, INSUFFICIENT DATA or NOT SUPPORTED. Nothing is sampled, interpolated or invented.

SOURCES

WHAT IS READ

NODE (JSON-RPC, SERVER-SIDE)
eth_getBlockByNumber(latest) → block height, head timestamp · eth_getCode → CONTRACT / EOA, bytecode size · eth_getTransactionCount → nonce · eth_getBalance → balance. All account reads are pinned to the same head block.
INDEX (BLOCKSCOUT V2, SERVER-SIDE)
The newest 100 transactions touching the address (sent or received), each with block, timestamp, status, fee, gas used and method. This is the SAMPLE every derived metric is computed from. The index is labelled on screen; if it does not answer, the profile says INDEXER UNAVAILABLE instead of showing zeros.

NETWORK REGISTRY

  • ROBINHOOD CHAIN · 4663RPC PUBLIC · BLOCKSCOUT
  • ETHEREUM · 1RPC PUBLIC · BLOCKSCOUT
  • BASE · 8453RPC PUBLIC · BLOCKSCOUT
  • ARBITRUM ONE · 42161RPC PUBLIC · BLOCKSCOUT
  • SOLANARECOGNISED, NOT READ. NO ADAPTER IMPLEMENTED YET.NOT SUPPORTED

EVM CHAINS SHARE ONE ADAPTER (JSON-RPC + BLOCKSCOUT). A NEW EVM NETWORK IS ONE DEFINITION FILE. SOLANA NEEDS ITS OWN ADAPTER AND HAS NONE.

METHODOLOGY

HOW EACH METRIC IS CALCULATED

ACTIVITY RATE

Number of sampled transactions whose timestamp is within the last 24 hours. Also shown as IN WINDOW 24H.

SUCCESS RATE / FAIL RATE

successful ÷ (successful + failed) and failed ÷ (successful + failed). Transactions whose status the index has not finalised are counted separately as UNKNOWN and excluded from both ratios.

AVERAGE EXECUTION COST

Arithmetic mean of the fee (in native wei) across sampled transactions that report a fee. Printed in the chain's native unit. AVG GAS USED is the mean of gas used — compute units.

FRESHNESS

Now − newest sampled transaction timestamp, in seconds. HEAD AGE is now − latest block timestamp and tells you how stale the node's view is.

MEDIAN GAP

Median of the time differences between consecutive sampled transactions. Needs at least two timestamps, otherwise UNAVAILABLE.

IN / OUT · COUNTERPARTIES

Direction is derived from whether the inspected address is the sender or recipient. Counterparties are the distinct other addresses in the sample.

VGEO LOAD SCORE · 0–100

LOAD = 40% ACTIVITY + 25% FREQUENCY + 15% COST + 10% SUCCESS + 10% FRESHNESS

  1. ACTIVITY = min(1, transactions in last 24 h ÷ 50).
  2. FREQUENCY = 1 when the median gap ≤ 60 s, 0 when ≥ 24 h, log-linear between.
  3. COST = log(avg gas used ÷ 21,000) ÷ log(2,000,000 ÷ 21,000), clamped 0–1. Heavier execution scores higher.
  4. SUCCESS = success rate (0–1).
  5. FRESHNESS = 1 when the newest transaction is ≤ 60 s old, 0 when ≥ 7 d, log-linear between.
  6. A component that cannot be computed contributes 0 and is shown as N/A in the breakdown. The weighted sum is scaled to 0–100 and rounded.

024

IDLE

2549

LOW

5074

ACTIVE

7589

HOT

90100

SATURATED

VGEOCPU SCORES ARE INTERNAL ANALYTICAL SIGNALS, NOT FINANCIAL RATINGS. THE FORMULA IS NOT AN INDUSTRY STANDARD; IT IS PRINTED HERE SO ANY NUMBER ON SCREEN CAN BE RECOMPUTED FROM THE RAW JSON.

LIMITS

WHAT IT DOES NOT DO

  • It is read-only. No wallet, no signing, no private keys anywhere.
  • The sample is the newest 100 transactions. Older history is not fetched; SAMPLE · MORE EXIST tells you when it is truncated.
  • Internal transactions and token transfers are not part of the sample yet.
  • Results are cached for 60 seconds per address and network.
  • Solana addresses are recognised and refused. Nothing is estimated for them.