CKB Pulse
Lessons

What CKB learned

Start here · Glossary · Test yourself · Feed

“Other”

55 lessons found. All lessons

Page 3 of 5

ckb-vm-js 0.2.0 (npm) checks header hash instead of rejecting the fieldWorkingChecked by the owner

With ckb-vm-js 0.2.0 from npm, verifyScriptGroup accepts an RPC header in context.tip_header that carries its own hash only when that hash matches the recomputed one, and fails with "header hash does not match" otherwise (the earlier vendored build rejected any `hash` field as unknown). Dropping `hash` before passing RPC headers or transactions remains safe because the verifier recomputes it.

No existing lesson covers how ckb-vm-js treats a hash field on RPC headers.

#Testing and devnet #Data encoding (molecule, hex, bigints) · Sources (1) · 2026-11-17

Mock syscalls must write back output lengthAdvancedFrom one fix

When a ckb-vm mock or fuzzing syscall layer implements inherited_fds, it must store the fd count back into length_addr with store64 after writing the fd buffer. Otherwise the script reads back the capacity it passed in and treats unwritten slots as fds. ckb-vm-contrib's fuzzing utils lacked this until v1.1.2.

#Testing and devnet · Sources (1) · 2026-11-16

Lock-based fallback when a stored out point goes staleWorkingFrom one fix

A saved out point stops being valid once the cell is consumed and recreated. Use the recorded out point if it is still live; otherwise find the unique live cell matching the exact lock (code_hash, hash_type, args). In both cases, still check the cell's contents against the recorded values.

#Indexer queries and cell collection · Sources (1) · 2026-11-15

Use findCellsOnChain for shared live stateWorkingFrom one fix

client.findCells merges results from the client's local cache, which can still hold outputs of your own earlier transactions after other parties spent them on chain. When you need the true live cell set, such as state shared with other actors or runners, use client.findCellsOnChain.

Medium confidence.

#Indexer queries and cell collection · Sources (1) · 2026-11-15

Sign a prepared transaction without broadcasting via CCC signOnlyTransactionWorkingFrom one fix

signer.signOnlyTransaction(tx) asks the CCC signer to sign a transaction you already completed (inputs, fee, cell deps, zero placeholder witness) without preparing or sending it. Treat the wallet as untrusted: pass it a copy, recheck the connected account after it returns, and verify the signed result against the sign hash you got from tx.getSignHashInfo(lock, client) on the prepared transaction (same tx hash, other witnesses untouched, signature recovers to the expected public key).

#Witnesses and signatures #Transaction building and fees · Sources (1) · 2026-11-09

CKB log subscription notifications may carry params.result as a JSON stringWorkingFrom one fix

When subscribing to the CKB node's "log" topic over the TCP subscription RPC, the notification's params.result can be a JSON-encoded string of the log entry (message, level, target, date) rather than an object. Parse it with JSON.parse when it is a string, and skip malformed payloads instead of stopping the stream; otherwise script debug output (target ckb_script::verify) is silently dropped.

#Network and RPC reliability #Testing and devnet · Sources (1) · 2026-11-09

Cheque lock sender withdrawal requires exact since 0xA000000000000006WorkingFrom one fix

The deployed Cheque lock (RFC 0039) accepts a sender withdrawal only when the Cheque input's since is exactly 0xA000000000000006 (relative epoch, 6 epochs after the cell was committed). A since of 0 fails with Cheque error 11 (WithdrawChequeInputSinceError), so the sender cannot reclaim for about 24h after each create or renew.

#Since and time locks #Testing and devnet · Sources (1) · 2026-11-09

Classifying sendTransaction failures before dropping a pending CKB transactionWorkingFrom one fix

A failed or timed out ccc client.sendTransaction, or a getTransaction status of "rejected" or "unknown", does not prove the transaction can never commit: a node may have taken it, and "rejected" also covers pool removals such as expiry or a full pool. Store the signed bytes before sending. Drop the pending state only on final refusals (ErrorClientVerification, ErrorClientMaxFeeRateExceeded, TransactionFailedToVerify, PoolRejectedMalformedTransaction), or when an input is not live, the transaction that created that input is committed, and the node no longer holds your transaction as pending, proposed or committed. A not-live input alone is not proof, because CCC's getCellLive also returns not live for a cell the node has not seen. Otherwise resend the stored bytes, treating ErrorClientDuplicatedTransaction and PoolRejectedDuplicatedTransaction as success.

#Network and RPC reliability #Transaction building and fees · Sources (1) · 2026-11-08

Iterating group inputs and outputs with ckb-std until IndexOutOfBoundWorkingFrom one fix

In a ckb-std script, walk the cells of a source with the high-level loaders, for example load_cell_type_hash(i, Source::GroupInput) or load_cell_lock_hash(j, Source::Output), in an index loop. Break on Err(SysError::IndexOutOfBound), and return a script error code for any other SysError. Accumulate the u64 capacities you load in a u128 so that a sum over many cells cannot overflow.

#Capacity and cell size #Cell deps and script lookup · Sources (1) · 2026-11-08

Multisig sighash placeholder lock: script plus M zeroed signaturesWorkingFrom one fix

When computing the signing message for a secp256k1_blake160_multisig_all input, set the first witness's WitnessArgs.lock to the multisig script bytes followed by threshold × 65 zero bytes (a single key sighash lock uses just 65 zero bytes), and keep the inputType and outputType fields, before calling CCC's tx.getSignHashInfo(script, client). The same placeholder length must be used here as in the signed witness, otherwise the message differs from what the lock script hashes on chain.

#Witnesses and signatures #Transaction building and fees · Sources (2) · 2026-11-08

CCC EVM signer exposes several lock addresses via getAddressObjsWorkingChecked by the owner

A CCC signer for MetaMask can control more than one CKB lock (Omnilock addresses and the legacy PW-lock address used by Portal Wallet), and getRecommendedAddressObj returns only one of them. When assets such as Nervos DAO deposits may sit under any of them, call signer.getAddressObjs(), dedupe by script hash and check each, rather than assuming the recommended address holds the funds.

#Indexer queries and cell collection #Login and authorization · Sources (1) · 2026-11-06

Locate the signing witness slot instead of hardcoding its indexWorkingFrom one fix

The signing witness of a lock group sits at the index of the first input that uses that lock, so it differs between transaction shapes (in QuotaCell, 1 for payments and most owner actions, 0 for a deposit). In CCC, take the position from tx.getSignHashInfo(lock, client).position and assert it matches the position you set, instead of hardcoding 1. A wallet or test mock that has no lock available can look for the placeholder witness, but a bare "first non-empty witness" search is only safe when no other input carries a witness.

#Witnesses and signatures · Sources (1) · 2026-11-05

Page 3 of 5

Text under CC BY 4.0: credit LusoCryptoLabs, CKB lessons, with a link to the lesson. JSON · Want a lesson changed or removed? Open an issue.