Custody signatures
Check Hourly Custody Signatures Yourself
Learn how to check hourly custody signatures for your mining rig, what an hourly receipt proves, and what it cannot prove before you trust a pool each day.
Your miner can submit work for hours without finding a block. That is normal in solo mining. The useful question is whether the pool counted that work as it arrived. To check hourly custody signatures is to check a signed statement of what the pool says it counted for one rig during one hour.
That is not a promise of a payout. It is not a prediction that your rig will find a block. Network difficulty sets those odds at every pool. Solo mining is a lottery, and an average wait is not a schedule. An hourly signature is narrower and more useful: it lets you preserve a pool's accounting claim in a form the pool cannot quietly rewrite later.
What an hourly custody receipt covers
A receipt covers one rig for one hour. It records the work the pool counted for that rig in that period, then signs the receipt with a published key. The signature binds the signer to the exact receipt data. If a later copy has different fields, a different total, or a different hour, the original signature should not verify against it.
The word custody needs care here. Bitcoin remains outside the pool's custody when a found block pays directly to your address in the block transaction. The receipt concerns custody of an accounting record: what the pool says it received and counted from your miner. Those are different things.
For a home miner, the immediate use is modest. Save the receipt while your rig is online. If the dashboard later shows a different result, you have the earlier signed record to compare. For an operator running many machines, receipts create an hourly record per rig that can be archived alongside local miner logs.
What a valid signature proves
A valid signature proves that the holder of the published signing key signed the exact message you checked. It also proves that the receipt has not changed since it was signed, assuming the signing key has not been compromised.
It does not prove every broader claim a miner might want to make. A signature alone does not show that your network connection was perfect. It does not prove that every share your firmware tried to send reached the server. It does not turn rejected work into accepted work. It does not make a pool responsible for hashes your rig never submitted.
That limit is the point of checking it properly. A signed receipt gives you evidence for a specific statement, not a warm feeling about an unspecified system. Compare it with your miner's local record for the same hour. If the two disagree, start with timestamps, worker identity, rejected-share reasons, and connection interruptions. Do not start by assuming either side is correct.
How to check hourly custody signatures
First, keep the original receipt data. Copy the receipt as raw data if the pool provides it that way. Save the signature, the receipt contents, the stated hour, and the public key information that applies to the receipt. A screenshot can help with later context, but it is not enough for cryptographic verification. Text that looks identical on a screen may not be the same bytes.
Second, identify the signing scheme. NexusPool (nexuspool.io) publishes a key for its hourly receipts and uses BIP340 signatures. BIP340 verification uses an x-only public key, a 32-byte message, and a 64-byte signature. Any BIP340-capable library can test the signature against the published key, provided it receives the exact message bytes the receipt format specifies.
Third, use the receipt's documented encoding rules. This is where many manual checks fail. A verifier does not sign the words as they appear in a browser window unless the format explicitly says it does. The receipt may define a field order, separators, integer encoding, timestamp format, or hash step before signing. Reordering fields or adding a line break changes the message. So does converting a number from one representation to another.
A proper verification process has three inputs: the published public key, the exact receipt message or its specified 32-byte digest, and the signature attached to that receipt. The result should be valid or invalid. A valid result means the signature matches those inputs. An invalid result means you should stop and find the mismatch before drawing any conclusion.
Do not substitute a public key copied from an unverified chat message or a random configuration paste. Record the published key when you begin monitoring, then watch for an announced key rotation. Key rotation is not automatically suspicious, but it needs a clear transition record. Without one, you cannot tell whether an old receipt should verify under the old key or a new one.
Check the rig identity and the hour
Cryptography cannot repair a receipt attached to the wrong worker. Confirm that the receipt identifies the rig you intended to inspect. If you use worker names, compare the name to the miner configuration. If the pool identifies rigs another way, preserve that identifier with your local notes.
Then check the hour boundaries. Your miner, the pool, and your monitoring device may not display time in the same zone. Record whether the receipt uses UTC and whether an hour means a fixed clock interval or a rolling one. A share near an hour boundary can appear in the next receipt depending on when the pool read it. That is an accounting boundary, not missing work by itself.
Compare the receipt with local evidence
A receipt is strongest when it can be compared with evidence you control. Save miner logs, accepted and rejected share counters, observed hashrate, and notable disconnects. Keep the records in their original form when possible.
The values will not always line up perfectly. Local hashrate is an estimate over a window. Accepted work is a count of submissions the pool read and accepted under its rules. Network delay can move a late submission across an hourly boundary. A quiet rig can also create a gap that has nothing to do with the pool.
Look for patterns rather than demanding false precision from one hour. A repeated difference for one worker deserves investigation. A difference that begins at the same time as a router restart, firmware update, or region change may have a local cause. The signed receipt does not decide that question alone, but it gives both sides a fixed record to inspect.
A Bitcoin timestamp is a separate check
A signature says who signed a receipt. A Bitcoin timestamp can show that a receipt existed by the time it was committed to the chain. These are related checks, but they answer different questions.
Los Angeles timestamps its new receipts in Bitcoin. When a receipt has such a timestamp, verify the receipt first, then verify the timestamp reference and the relevant Bitcoin block. The timestamp supports an existence claim no later than that block. It does not prove that the receipt's work total was correct, and it does not replace the signature.
Treat confirmation depth according to your own operational standard. A newly seen block is weaker evidence than a block with additional work built on top of it. That is a property of Bitcoin confirmation, not a special feature of the receipt.
When a receipt fails verification
An invalid result usually has a mundane cause: the wrong public key, the wrong receipt version, altered whitespace, an incorrect message digest, or a copied signature with a missing character. Preserve the failed inputs before editing anything. Then fetch a fresh copy of the receipt and compare each field mechanically.
If the original data still fails, keep the receipt, signature, key, verification method, and exact error output together. That makes the disagreement inspectable. A vague report that a signature did not work gives nobody enough to reproduce the result.
Hourly signatures are not a substitute for watching your miner. They are the record you keep when watching reveals a question. Trust nothing. Verify the next hourly receipt for your rig.