Every captured commit receives deterministic security triage and a separate communication-quality score. Security candidates and broader second-pass signals receive full-patch Ollama analysis.
Message quality measures whether a commit identifies its scope, purpose, rationale, testing, and supporting references. It does not change the security-severity score.
This is a large internal code reorganization (refactor) in Bitcoin Core. It creates a new BlockTemplateManager class that takes over block-template creation, block submission, and tip-waiting helpers that were previously spread across seve…
Large refactor touching mining, RPC, interfaces, and test shutdown pathsNew object lifetime dependency: BlockTemplateManager holds references to mempool, chainman, and notifications; explicit reset ordering added in Shutdown/InitAndLoadChainstate/test setupsRemoval of early-init node.mining interface; BlockTemplateManager is now created after chainstate load, with a comment that it must exist before setChainstateLoaded(true) unblocks IPC waiters
This commit adds the first implementation of BIP352 (Silent Payments) to Bitcoin Core. Silent Payments are a new type of privacy-preserving Bitcoin address that lets someone receive payments without publicly revealing a fixed address. The …
New cryptographic feature implementation (BIP352 Silent Payments)Extensive use of secp256k1 silentpayments moduleInput public key extraction from P2PKH, P2WPKH, P2SH-P2WPKH, and P2TR inputs
This update fixes a wallet database loading bug where a damaged or tampered Bitcoin wallet file could cause the program to read past the end of a stored extended public key (xpub). The patch makes the loader check the stored xpub length be…
Out-of-bounds read in wallet descriptor cache deserializationASan container-overflow triggered by malformed on-disk recordMissing length validation between record size prefix and fixed-size decoder
This commit adds a new wallet RPC called listrawtransactions to Bitcoin Core. It is a feature addition that lets users list every transaction their wallet knows about, including internal transfers and consolidations that the existing listt…
No security-relevant bug fix or vulnerability patch is present in the diff.New RPC exposes additional wallet transaction metadata, but only to callers already authorized for wallet RPCs.Code is a refactor of existing gettransaction logic into shared helpers; no new cryptographic, network, or consensus code.
This Bitcoin Core update fixes several wallet bugs where a failed database write could leave a wallet in an inconsistent state. For example, encrypting a wallet or changing its passphrase could appear to succeed in memory while the change …
Atomicity fix for encryption state and descriptor key persistenceFailure to persist master key during encryption previously reported success in memoryPassphrase change could activate new passphrase only in memory
This commit only changes Bitcoin Core's internal functional test code. It replaces hard-coded test keys and addresses with ones generated from a new test helper class, and unifies how tests tell nodes not to create a default wallet. There …
This commit only adds a new automated test to Bitcoin Core. It checks that when two partially-signed Bitcoin transactions (PSBTs) are combined, any custom 'unknown' data fields attached to them are preserved correctly. There is no change t…
This is a Bitcoin Core wallet maintenance patch. It speeds up a wallet function that checks whether a descriptor already exists by caching a hash of the descriptor's canonical text, instead of rebuilding that text every time. It also tidie…
No security-relevant signal in commit message or diffChange is described as performance improvement and code cleanupBackwards-compatibility test notes a known miniscript wallet loading incompatibility between v31.0/v31.1 and other versions, but this is a documented compatibility quirk, not a vulnerability
This is a documentation-only fix for Bitcoin Core's machine-readable RPC help data. It changes several default values from literal strings to 'hint' labels (because the real default depends on context) and corrects one boolean default from…
OpenRPC schema/default mismatch correctionRPC help metadata type correction (string 'false' to boolean false)No executable code path changes
This commit fixes documentation metadata for six Bitcoin Core RPC arguments. It changes how default values are described so that automatically generated API docs and schemas are accurate. The actual behavior of the software when running is…
No runtime code changesOnly RPC help/schema metadata modifiedVendor explicitly states runtime behavior is unchanged
This commit fixes a bug in Bitcoin Core's MuHash3072 cryptographic code where dividing a MuHash object by itself (x /= x) produced the wrong mathematical result. The fix is straightforward: the code now saves the divisor's numerator before…
Cryptographic correctness bug in MuHash3072 division operatorSelf-aliasing in operator/= produces incorrect 1/D result instead of empty setNo production code path identified that triggers self-division
This is a build-compatibility fix, not a security patch. It changes how some constant data is stored internally so that Apple's macOS linker (ld64) can build Bitcoin Core correctly. The change avoids a linker bug that caused build failures…
No security-relevant code logic changedChange is a linker bug workaround, not a vulnerability fixConstants remain read-only; no new attack surface introduced
This commit refactors Bitcoin Core's wallet descriptor import feature so the same logic can be used by both the RPC command and a new GUI-facing interface. It also tightens one input rule: negative timestamps are now rejected, and the mini…
Refactor of security-sensitive wallet import code into shared CWallet pathNew input validation: negative timestamps rejected for importdescriptorsCentralization of descriptor range bound checks in CheckDescriptorRangeBounds
This change fixes a bug in how Bitcoin Core reconnects to the Tor control port. A previous update accidentally removed the wait time between reconnect attempts when an already-established Tor control connection was dropped. Without the wai…
Uncontrolled retry loop causing resource exhaustion and log floodingLocal-only Tor control port interaction; no remote attacker path by defaultRegression introduced by prior refactor (#34158) and restored here
This change updates the Windows code-signing tool used in Bitcoin Core's reproducible build process. It fixes a build-time failure where signature verification could not complete because a certificate package was missing and the old tool v…
Tooling update in release signing pipelineRestores CA certificate store for signature verificationDisables CRL/CDP network lookups during verification
This change lets Bitcoin Core store different custom signet blockchains in separate data folders, using a unique suffix derived from each signet's network identifier. It also adds a friendlier error hint in bitcoin-cli when an RPC authenti…
Data isolation between distinct custom signets reduces risk of cross-network state corruption or accidental mainnet/testnet confusionNo memory-safety, cryptographic, or consensus changes observedNo privilege escalation, remote code execution, or denial-of-service vectors introduced in the diff
This change adds two extra pieces of information—whether a spent output came from a coinbase transaction and the block height at which it was created—to a Bitcoin Core REST API endpoint. It is a feature/parity improvement to make the REST …
This change makes Bitcoin Core treat manually-added peers (from -addnode, -connect, or the addnode RPC) more gently during Initial Block Download (IBD). Previously, if such a peer was slow or stalled at sending blocks, the node would disco…
Behavior change in peer disconnection logic during IBDManual peers exempted from block-stalling disconnectionNew per-peer cooldown state m_block_download_paused_until introduced
This is a small fix in Bitcoin Core's own test helper code. A helper function used only in tests could crash with an IndexError when given an extremely short fake signature, instead of cleanly returning False. The change moves a length che…
Out-of-order bounds check leading to IndexError in test helperRegression test added for malformed short DER signaturesTest-only code path, no production validation logic changed
This commit only adds a new functional test to Bitcoin Core. It checks that the getrawtransaction RPC can retrieve a stale block's coinbase transaction via the optional txindex, and that the response correctly shows the block is no longer …
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides an explanatory body
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This commit renames and expands a command-line helper in Bitcoin Core. The old 'bitcoin-util netmagic' command, which only printed the network's magic bytes, is replaced by 'bitcoin-util getchainparams', which prints a JSON document containing chain parameters such as genesis block, ports, DNS seeds, and address prefixes. There is no security fix or vulnerability here; it is a routine feature change.
AI review queuedwallet: Show alternate wtxids in gettransactionby Ava Chow · 2d55c7a7 · Jul 16, 2026 · 3 filesMessage 60 · AdequateInformational 18Details
Commit message · Ava Chow
wallet: Show alternate wtxids in gettransaction
If a wallet tranasction has alterate witness versions, list those wtxids in gettransaction's output.
60/100 · AdequateMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Provides an explanatory body
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 18/100
This commit adds a new field called 'alternate_wtxids' to the output of the wallet RPC command 'gettransaction' (and related transaction listings). It simply shows other witness transaction IDs that share the same base transaction ID. This is an informational UI/API change, not a fix for a vulnerability.
AI review queuedwallet: Replace CWalletTx::SetTx with Updateby Ava Chow · 0b1af01b · Jul 16, 2026 · 3 filesMessage 80 · StrongLow 42Details
Commit message · Ava Chow
wallet: Replace CWalletTx::SetTx with Update
Instead of replacing the tx when a witness alternative appears, add it to the set of wtxid alternates.
In order to determine whether the added transaction is the canonical transaction, Update also needs to know how the state is changing, so it will also update the state if it is being changed.
80/100 · StrongMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Explains rationale or failure mode
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Low 42/100
This Bitcoin Core wallet commit changes how the wallet stores transactions that share the same transaction ID (txid) but have different witness data (wtxid). Previously, the wallet would overwrite the stored transaction when a version with witness data appeared. Now it keeps all known variants and picks a 'canonical' one based on rules: confirmed transactions win; otherwise, transactions with witness data are preferred, and lighter-weight transactions beat heavier ones. The change is a defensive refactor to support multiple transaction variants correctly, not a fix for a known active exploit.
AI review queuedtest: Test for wallet txs with alternate wtxidsby Ava Chow · ef2afc6a · Jul 16, 2026 · 1 fileMessage 67 · AdequateInformational 14Details
Commit message · Ava Chow
test: Test for wallet txs with alternate wtxids
67/100 · AdequateMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Uses a recognizable type or scope✓ Mentions testing or verification! No meaningful explanatory body
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 14/100
This commit adds a new automated test for Bitcoin Core's wallet. It checks that when the same transaction can be spent in two different ways (a lighter 'key path' and a heavier 'script path' on a Taproot output), the wallet correctly reports which version is 'canonical' and lists the other versions as 'alternates'. The commit only adds test code; it does not change wallet logic or fix a bug.
AI review queuedwallet: Store all witness variants of a transactionby furszy · 56cf27db · Jul 16, 2026 · 5 filesMessage 85 · StrongLow 35Details
Commit message · furszy
wallet: Store all witness variants of a transaction
A transaction can have several valid witnesses that share its txid but differ in wtxid, e.g. when a taproot output is spent via the key path in one variant and the script path in another.
CWalletTx now keeps all of them in a map indexed by wtxid (m_txs) and marks one as canonical (m_canonical_wtxid): a confirmed variant if there is one, otherwise the one with a witness and the lowest weight. GetTx() and serialization return the canonical variant, so existing callers don't need to change.
The other variants are stored in their own wtxvariant records keyed by (txid, wtxid) and merged back into the CWalletTx at load. The tx record keeps its old format, holding the canonical transaction, so older soft versions can still read and rewrite it without dropping those records.
85/100 · StrongMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Explains rationale or failure mode
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Low 35/100
This Bitcoin Core wallet change lets the wallet remember multiple valid versions of the same transaction that look identical on-chain by transaction ID (txid) but differ by witness transaction ID (wtxid). For example, a Taproot spend can be signed in two different valid ways. Previously the wallet only kept one witness version, which could cause it to lose track of the actual confirmed version or show wrong balances/labels. The fix stores all variants and picks a canonical one, with older wallet software still able to read the main record safely.
AI review queuedwallet: Remove unused CWalletTx CopyFrom and copy constructorby Ava Chow · 72ebdd63 · Jul 16, 2026 · 2 filesMessage 50 · ThinInformational 22Details
Commit message · Ava Chow
wallet: Remove unused CWalletTx CopyFrom and copy constructor
50/100 · ThinMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component! No meaningful explanatory body
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 22/100
This commit removes an unused explicit copy helper and strengthens a safety guard in Bitcoin Core's wallet code. Previously, copying a wallet transaction object was allowed through a private default copy constructor and an explicit CopyFrom() method. The change deletes both, so the object can no longer be accidentally duplicated. The commit message and code comment explain this is meant to prevent bugs where updates happen on the wrong copy of a transaction, not to fix an active security vulnerability.
AI review queuedwallet: Make CWalletTx::tx private and use CWalletTx::GetTx to accessby Ava Chow · 798ba6d0 · Jul 16, 2026 · 14 filesMessage 73 · AdequateInformational 19Details
Commit message · Ava Chow
wallet: Make CWalletTx::tx private and use CWalletTx::GetTx to access
When CWalletTx will have multiple transactions, tx will no longer exist and accessing the single canonical tx should be done through an getter function.
73/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 19/100
This is a routine internal code cleanup in Bitcoin Core's wallet module. It makes the transaction pointer inside a wallet transaction object private and forces the rest of the code to read it through a getter function. There is no security fix here; the change prepares the code for a future feature where a wallet transaction might hold more than one transaction.
AI review queuedwallet: Deserialize directly in CWalletTx's ctorby Ava Chow · 19af439b · Jul 16, 2026 · 5 filesMessage 80 · StrongLow 45Details
Commit message · Ava Chow
wallet: Deserialize directly in CWalletTx's ctor
When loading a transaction, instead of constructing a CWalletTx with no transaction, pass the DataStream into the constructor so that the CWalletTx is RAII. This allows us to ensure that the transaction is never a nullptr so that dereferences, especially once multiple txs are stored, will not cause a segfault.
80/100 · StrongMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Explains rationale or failure mode
Why it was queued
signing or wallet pathsecond-pass: broader security terminologysecond-pass: security-sensitive path
AI analysis · Low 45/100
This commit changes how Bitcoin Core wallet transactions are loaded from disk or copied between wallets. Previously, a wallet transaction object could be created with no actual transaction inside it, and later filled in. The patch makes the transaction data mandatory from the start using a constructor that deserializes directly from a data stream. The stated goal is to prevent crashes (segfaults) caused by accidentally using a null transaction pointer, especially as more transaction types are stored in the wallet. It also improves error handling for corrupt wallet records.
This overload has been unused since it was added in fac81affb527132945773a5315bd27fec61ec52f.
75/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides an explanatory body✓ Mentions testing or verification
Why it was queued
fuzzing or regression evidencesecond-pass: broader security terminology
AI analysis · Informational 15/100
This commit simply removes an unused helper function inside Bitcoin Core's fuzz testing code. Fuzz tests are internal developer tools used to find bugs, not code that runs on live Bitcoin nodes or wallets. There is no security-relevant change to production behavior.
AI review queuedipc, refactor: Change Protocol class field orderby Ryan Ofsky · 00287b9a · Jul 14, 2026 · 1 fileMessage 68 · AdequateLow 26Details
Commit message · Ryan Ofsky
ipc, refactor: Change Protocol class field order
This just changes Protocol class field order to make sure class members are not destroyed before the event loop thread exits. There is no change in behavior. The change is just being made to clarify intent and avoid potential bugs.
68/100 · AdequateMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
parser or protocol pathsecond-pass: security-sensitive path
AI analysis · Low 26/100
This is a small defensive code change in Bitcoin Core's inter-process communication (IPC) code. It reorders two class member variables so that the event loop is destroyed before the background thread that runs it. The commit message says there is no behavior change and the goal is to clarify intent and avoid potential future bugs. The change itself does not fix a known exploitable vulnerability, but it addresses a real C++ object lifetime risk.
AI review queuedipc, refactor: Add SocketId type alias and use itby Ryan Ofsky · 3859805f · Jul 14, 2026 · 7 filesMessage 68 · AdequateInformational 15Details
Commit message · Ryan Ofsky
ipc, refactor: Add SocketId type alias and use it
Use SocketId type instead of int to represent socket ids to be compatible with an upcoming version of libmultiprocess which adds windows support.
68/100 · AdequateMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
parser or protocol pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This commit is a straightforward code cleanup: it replaces the plain 'int' type with a new 'SocketId' type alias when referring to socket identifiers throughout Bitcoin Core's inter-process communication (IPC) code. The change is described by the author as preparation for a future version of the libmultiprocess library that will add Windows support. There is no security fix here and no change to program logic—only type names and a constant for the error value were updated.
AI review queuedipc, refactor: Add Stream type alias and use itby Ryan Ofsky · e9f19815 · Jul 14, 2026 · 5 filesMessage 68 · AdequateInformational 15Details
Commit message · Ryan Ofsky
ipc, refactor: Add Stream type alias and use it
Use Stream type to abstract socket ids and be compatible with updated mp::ConnectStream() and mp::ServeStream() functions that use streams instead of socket ids in an upcoming version of libmultiprocess which adds windows support.
Since creating Stream objects from socket ids can require the event loop to be running, the ipc::Protocol::serve() method is also updated to accept the server stream though a callback parameter instead of a normal parameter.
68/100 · AdequateMessage clarity
✓ Descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
parser or protocol pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This is a straightforward internal code cleanup in Bitcoin Core's inter-process communication (IPC) layer. It replaces direct use of raw socket identifiers with a new 'Stream' type alias so the code can work with an upcoming version of a supporting library that adds Windows support. No security bug is fixed or introduced in the visible changes.
AI review queuedipc, refactor: Drop connect/listen/serve exe_name parametersby Ryan Ofsky · 33d37f3c · Jul 14, 2026 · 5 filesMessage 73 · AdequateInformational 15Details
Commit message · Ryan Ofsky
ipc, refactor: Drop connect/listen/serve exe_name parameters
Pass exe_name parameter to ipc::Protocol class constructor instead. It never really made sense to have exe parameters as part of the protocol interface and removing them makes adding new features like windows support easier.
The exe name values are only used for logging and debuggging purposes to distinguish log messages from different processes.
73/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
parser or protocol pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This is a straightforward internal code cleanup in Bitcoin Core's inter-process communication (IPC) layer. It moves the executable name parameter from individual connect/listen/serve methods into the protocol object's constructor. The executable name is only used for logging and debugging to tell different processes apart. There is no change to security-sensitive behavior, no bug fix, and no externally reported issue.
clusterlin: avoid heap allocations in GetLinearization
Replace GetLinearization's four per-call heap-allocated vectors (the ready chunk/transaction heaps and both dependency counters) with stack arrays and explicit size counters. This especially benefits linearization of small clusters, where the allocations make up a significant fraction of the total runtime.
73/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
second-pass: broader security terminology
AI analysis · Informational 15/100
This is a routine performance optimization in Bitcoin Core's transaction-cluster linearization code. It replaces dynamically-allocated vectors with fixed-size stack arrays to avoid repeated heap memory allocations. There is no security-relevant change: the algorithm, ordering logic, and dependency handling remain identical.
clusterlin: avoid recomputing intersections in MergeChunks
MergeChunks scans the top chunk's transactions twice: once to count the dependencies on the bottom chunk, and once to locate the randomly picked one. Remember the per-transaction dependency counts from the first pass, so the second pass only computes the intersection of the selected transaction, rather than one per transaction scanned.
73/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
second-pass: broader security terminology
AI analysis · Informational 15/100
This is a small internal performance improvement in Bitcoin Core's transaction clustering code. It avoids doing the same bit-set intersection twice by saving the result of the first pass. There is no security-relevant change: no new behavior, no bounds check changes, no memory safety changes, and no externally reachable input handling.
AI review queuedfuzz: Use LIMITED_WHILE over for-loop with consumed size integralby MarcoFalke · fa55385a · Jul 14, 2026 · 1 fileMessage 83 · StrongInformational 15Details
Commit message · MarcoFalke
fuzz: Use LIMITED_WHILE over for-loop with consumed size integral
This is a style cleanup. The general pattern to use `LIMITED_WHILE`, which all other fuzz tests use, has some benefits:
* When no data is available, a simple and single (let's say) 64 value in the fuzz input will not result in 64 loops over the same body with the same default/fallback values. * When no data is available, `ConsumeBool` falls back to `false` and breaks the loop early. * When further data is available, the overhead is just a single byte, making it also possibly easier for the fuzz engine to mutate the data, as a single int that influences the whole remainder of the fuzz input can lead to the 'havoc' effect. * When a crash is reduced, deleting bytes will directly influence the execution length, so byte-length of the fuzz input roughly corresponds to run-time length.
83/100 · StrongMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Mentions testing or verification
Why it was queued
fuzzing or regression evidencesecond-pass: broader security terminology
AI analysis · Informational 15/100
This commit is a minor style cleanup inside a fuzz test (a test that feeds random data to software to find bugs). It replaces a loop that reads a number from fuzz input with a standard helper that reads a single true/false byte each iteration. There is no change to production Bitcoin Core code, no user-facing behavior change, and no security fix.
Squashed 'src/ipc/libmultiprocess/' changes from 28e056576a..e8de5c7b68
e8de5c7b68 Merge bitcoin-core/libmultiprocess#305: refactor: memcpy to std::ranges::copy to work around ubsan warn 9307e68e5a Merge bitcoin-core/libmultiprocess#306: doc: Bump version 12 > 13 fac7b9b7f6 refactor: memcpy to std::ranges::copy to work around ubsan warn 1bd7025609 Merge bitcoin-core/libmultiprocess#297: test: add map serialization round-trip coverage 438fdd243d doc: Bump version 12 > 13 463d073cb8 test: rename vBool to vector_bool 85df233845 test: add mapStringInt to foo.capnp to cover map serialization and deserialization
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Mentions testing or verification✓ Links an issue, advisory, or supporting reference
Why it was queued
second-pass: broader security terminology
AI analysis · Informational 19/100
This commit updates a helper library (libmultiprocess) used by Bitcoin Core. The only code change replaces a raw memory copy with a safer standard-library copy in a routine that converts byte arrays for inter-process communication. The rest is version bookkeeping and new test coverage. There is no indication this fixes an active security vulnerability; it is a defensive cleanup to silence an undefined-behavior sanitizer warning.
✓ Subject identifies a change✓ Names a concrete action or component! No meaningful explanatory body
Why it was queued
second-pass: opaque commit message
AI analysis · Informational 15/100
This commit simply updates the version of a build dependency (Cap'n Proto, a data serialization library used during compilation) from 1.4.0 to 1.5.0 in Bitcoin Core's dependency system. It changes only a version number and the corresponding cryptographic hash used to verify the downloaded source. There is no indication in the commit itself of any security bug, vulnerability, or malicious change.
AI review queuedbench: replace CreateMockableWalletDatabase with MakeInMemoryWalletDatabaseby Pablo Martin · 7508ac31 · Jul 13, 2026 · 5 filesMessage 91 · StrongInformational 15Details
Commit message · Pablo Martin
bench: replace CreateMockableWalletDatabase with MakeInMemoryWalletDatabase
Benchmarks don't need mock-specific behaviour (overridden Filename(), Format(), or the exposed batch-level WriteKey()). Replace CreateMockableWalletDatabase() with MakeInMemoryWalletDatabase() across 6 call sites in src/bench/ (5 files), using the same in-memory SQLite path that production code uses.
wallet_migration.cpp is excluded: it calls GetOrCreateLegacyDataSPKM() which asserts Format() == "sqlite-mock", a signal used to allow legacy SPKM setup in test/bench contexts. MockableSQLiteDatabase is still correct there.
For coin_selection.cpp, which had no other dependencies on wallet/test/util.h, also switch the include to <wallet/sqlite.h>.
Follow-up suggested in #35655.
91/100 · StrongMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Mentions testing or verification✓ Links an issue, advisory, or supporting reference
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This commit is a small internal cleanup in Bitcoin Core's benchmark code. It swaps a test-only mock wallet database helper for a real in-memory SQLite database helper in five benchmark files. There is no change to production wallet code, no change to how real user wallets are stored, and no security fix or vulnerability introduced.
AI review queuednet_processing: Drop unnecessary txid arg from InitiateTxBroadcastToAllby Anthony Towns · 349c72ee · Jul 11, 2026 · 3 filesMessage 50 · ThinInformational 15Details
Commit message · Anthony Towns
net_processing: Drop unnecessary txid arg from InitiateTxBroadcastToAll
50/100 · ThinMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component! No meaningful explanatory body
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This is a small code cleanup change in Bitcoin Core's network transaction broadcasting. It removes an unused transaction ID argument from an internal function called InitiateTxBroadcastToAll, leaving only the witness transaction ID. The function's behavior is unchanged; only its signature and call sites are simplified.
wallet: Introduce WalletError with machine-readable error code
Introduce WalletError as a generic wallet-layer error type that can carry a machine-readable WalletErrorCode and a translated user-facing message.
The WalletErrorCode::GenericError code is intended for failures that callers should only display to the user. More specific codes should only be added when callers can handle the condition differently.
73/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 15/100
This commit only adds a new data structure for wallet error reporting. It does not change any behavior, fix a bug, or alter how funds or keys are handled. There is no security issue here.
✓ Descriptive subject! No meaningful explanatory body
Why it was queued
second-pass: opaque commit message
AI analysis · Informational 16/100
This commit adds an empty placeholder for NetBSD linker flags in the build system. It is a one-line build configuration change with no functional or security effect visible in the diff.
AI review queuedwallet: use in-memory SQLite for temporary wallet in exportwatchonlywalletby Pablo Martin · d1e7f8c9 · Jul 8, 2026 · 5 filesMessage 63 · AdequateInformational 18Details
Commit message · Pablo Martin
wallet: use in-memory SQLite for temporary wallet in exportwatchonlywallet
The intermediate watchonly wallet created during exportwatchonlywallet is a pure build artifact — it is always discarded once BackupWallet() copies it to the destination. Creating it as an in-memory SQLiteDatabase (SQLITE_OPEN_MEMORY) removes the need to write files to the wallets directory and eliminates the cleanup handler that deleted those files on both success and failure paths.
Introduces InMemoryWalletDatabase (a minimal SQLiteDatabase subclass) and MakeInMemoryWalletDatabase() factory in sqlite.h/cpp, following the same pattern as MockableSQLiteDatabase / CreateMockableWalletDatabase() in the test utilities. MockableSQLiteDatabase now derives from InMemoryWalletDatabase, removing its redundant Files() override.
The wallet is named after the source wallet ("<name>_watchonly_temp") so concurrent exports of different wallets use distinct names and log lines remain traceable to the source wallet.
63/100 · AdequateMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Mentions testing or verification! Contains work-in-progress language
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 18/100
This change refactors how Bitcoin Core creates a temporary wallet during the exportwatchonlywallet command. Previously, the temporary wallet was written to disk in the wallets directory and then deleted afterward. Now it is kept entirely in memory using SQLite's in-memory mode, so no temporary files are created on disk. This is a cleanup and hardening improvement, not a fix for an active exploit.
AI review queuedwallet: store m_additional_flags in SQLiteDatabase to fix reopen pathby Pablo Martin · ee43743f · Jul 8, 2026 · 2 filesMessage 73 · AdequateLow 31Details
Commit message · Pablo Martin
wallet: store m_additional_flags in SQLiteDatabase to fix reopen path
SQLiteDatabase::Open() (the public override) always reopens the database with no additional flags. If SQLiteBatch::Close() triggers the force_conn_refresh path (TxnAbort failed), it calls Open() which drops the original additional_flags, causing in-memory databases to be reopened as on-disk instead.
Store additional_flags as a member and use it in Open() so the reconnect preserves the original flags. For in-memory databases, connection recovery makes no sense as all data would be lost; both the force_conn_refresh path and the public Open() now throw instead.
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Low 31/100
This patch fixes a bug in Bitcoin Core's SQLite wallet code where an in-memory test wallet could accidentally be reopened as a real on-disk database, or be lost entirely during error recovery. The fix stores the original database flags so reopening preserves them, and explicitly throws an error instead of trying to recover a connection to an in-memory database (which would lose all data). It appears to be a correctness/reliability fix rather than an active security vulnerability in normal production use.
AI review queuedrpc: add `vsize_adjusted` field to getrawtransaction output for mempool transactionsby Musa Haruna · 5d25a0c2 · Jul 7, 2026 · 2 filesMessage 83 · StrongInformational 19Details
Commit message · Musa Haruna
rpc: add `vsize_adjusted` field to getrawtransaction output for mempool transactions
Extend the `getrawtransaction` RPC to include a new field `vsize_adjusted` when the transaction is in the mempool. The `vsize_adjusted` field provides the mempool's accounting size for the transaction based on its sigop cost, which can exceed its serialized vsize under `-bytespersigop` policies.
Test coverage is added to verify the correct calculation and exposure of the `vsize_adjusted` field via `mempool_sigoplimit.py`.
83/100 · StrongMessage clarity
✓ Specific, descriptive subject✓ Names a concrete action or component✓ Provides detailed explanatory context✓ Mentions testing or verification
Why it was queued
signing or wallet pathsecond-pass: security-sensitive path
AI analysis · Informational 19/100
This commit adds a new read-only field called vsize_adjusted to the getrawtransaction RPC output, but only when the transaction is currently in the memory pool. It exposes information the node already calculates for mempool policy, so it does not change behavior, fix a bug, or introduce a vulnerability. It is a transparency/feature improvement for RPC users.