Two causes of the "every message I send shows up again as a reply from the peer":
1. Harvest bug — the full-refresh path fed a SENT tx's outgoing memos (from
z_viewtransaction outgoing outputs) into the chat metadata extractor, and
ChatService::ingest marks everything Incoming. So each send was re-ingested as
a phantom "from peer" message. The recent-refresh path already omitted this, so
it was accidental. Drop the outgoing chat-harvest (keep the tx-history harvest);
genuine incoming still comes from z_listreceivedbyaddress, and our sends are
recorded by the local echo.
2. Own-identity ingest filter — a memo whose sender public key equals our own
identity is by definition something we sent (only we hold our key); it must
never be ingested as incoming. Skip those in ingest. This also collapses
same-seed self-chat (running the SAME wallet in the full node and Lite makes
them one chat identity, so sends land on an address we also own and loop back).
For a real two-party chat use two DIFFERENT wallets/seeds — same-seed wallets are
one identity and can't be distinct peers.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Four confirmed findings from the review of ef247c9:
1. Persistence regression — the deferred-persist echo (in-memory Sending, written
only when the async callback resolved) meant a message broadcast on-chain but
whose callback hadn't fired yet was LOST from history if the app quit/crashed
in that window. Persist the echo immediately as Sending and UPSERT the final
status on resolve (new ChatDatabase::upsert with ON CONFLICT DO UPDATE, since
append is INSERT-OR-IGNORE). A stray persisted Sending still loads as Sent.
2. Fee ceiling — dragonxd REJECTS a 0-value tx whose fee exceeds the default
miners fee (0.0001), and max(getDefaultFee(), 0.0001) can only raise it, so a
default_fee > 0.0001 broke every chat send. Pin chat to exactly kChatMinFeeDrgx,
dropping getDefaultFee() from this path (chat always moves 0 value).
3. Lifetime — the resolve callback had no generation guard, so a wallet lock (which
doesn't disconnect) between submit and callback could resolve against a cleared
store. Capture chat_session_generation_ and bail on mismatch (both the full-node
callback and the lite optimistic resolve), matching the identity-fetch pattern.
4. Retry misdirect — Retry on a failed CONTACT REQUEST called sendChatMessage,
which (no peer key yet) just showed "waiting for reply". Route it to
sendContactRequestForCid() (refactored out of startChatConversation) so it
re-sends the request into the SAME conversation.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Chat sends move 0 value, so the network fee is structurally load-bearing (it's
the only thing that forces a real shielded input; 0-value + 0-fee builds a
degenerate, unrelayable tx). Three gaps addressed:
1. Fee floor — broadcastChatMemos now uses max(getDefaultFee(), kChatMinFeeDrgx),
so a 0 / too-low global default-fee setting can't silently break chat.
2. Real delivery status — the echo was marked Sent on SUBMIT regardless of the
on-chain outcome (the z_sendmany callback was empty), so failures were
invisible and the Retry affordance never fired for async failures. Add a third
ChatDelivery::Sending state (appended so persisted 0=Sent stays valid); record
the echo in-memory as Sending, and resolve it to Sent/Failed from the
z_sendmany completion callback — persisting only the final status (so a restart
never shows a stuck spinner; a stray persisted Sending loads as Sent). A subtle
"sending…" label shows while in flight.
3. Pay-from-funded + pre-check — z_sendmany spends from one z-address, and the
identity reply address may be unfunded while funds sit elsewhere. chatPayFromZaddr
picks a spendable z-address that can cover the fee (preferring the identity
address); the memo still advertises the identity address as reply-to, so paying
from a different note is transport-transparent. If nothing can cover the fee, a
clear "need a small shielded balance" toast replaces the cryptic failure.
Full node only for the callback path; lite resolves optimistically on queue.
8-language strings + CJK subset (+1 glyph 賄).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Four confirmed findings from the review pass:
SECURITY — B7 was incomplete:
- The single-key Export dialog (key_export_dialog) still used plain call() for
z_exportkey/dumpprivkey/z_exportviewingkey — a live spending-key leak on the
most common per-address export path, missed by the B7 commit.
- callSecret() zeros the raw body but the parsed json holds its OWN heap copy of
the secret; several callers did .get<string>() on a temporary json and freed
that copy un-wiped.
Fix: add RPCClient::callSecretString() — returns the bare-string result with
BOTH the raw body AND the json node zeroed, so callers can't forget. Route
key_export_dialog (×2), exportPrivateKey, and export_all_keys (×2) through it;
scrub the z_exportmnemonic json node in seed_wallet_creator (object result);
also wipe the transient key copies, the displayed s_key on reset, and the
aggregated export-all `keys` buffer.
CHAT:
- Jump-to-latest pill: SetCursorScreenPos moved the parent cursor and never
restored it, so the composer footer rendered ~8px too high while scrolled up.
Save + restore the cursor around the pill.
- New-message toast: gating on a chatUnreadCount() watermark delta could be
swallowed when an outgoing echo (wall-clock) pushed the seen-watermark past a
later reply's block time. ingest() now reports the cids it appended; the toast
fires when any is a non-muted conversation — skew-proof, still mute-aware.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
B4: the legacy chat-identity fallback copied the z_exportkey spending key out of
the RPC response and let the temporary json destruct un-zeroed. Mirror the
mnemonic path — take our copy, then sodium_memzero the json's own buffer.
B2: the memo header's peer z-address / cid ride OUTSIDE the secretstream AEAD, so
trusting the newest message's header let a later message redirect our replies or
splice threads. Pin the reply target (and displayed peer) to the EARLIEST
(establishing) message instead of the latest, at both the send and display sites.
Because ChatStore returned filtered INSERTION order (a scan harvests txids in
set/hash order — not chronological), "earliest" wasn't reliable; ChatStore::
conversation now returns messages sorted by (timestamp, txid, payload_position),
which also fixes out-of-order thread rendering and the last-message preview.
A complete fix binds z+cid into the AEAD additional-data, but that's a coordinated
HushChat/SDXLite wire-format change; this pin hardens the reply target without it.
Adversarially verified; the store-ordering gap it surfaced is fixed here.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Outgoing echoes were recorded optimistically regardless of whether the broadcast
was actually submitted. Now broadcastChatMemos / broadcastChatMemosLite return
whether the send was submitted (false on immediate failures: not connected, no
spendable z-address, or a lite send already in progress), and the echo records a
ChatDelivery status (Sent / Failed). The Chat tab shows a "not sent" marker in
the error color on failed messages. The status persists in the DB (backward-
compatible: old rows deserialize as Sent).
(A later async failure — an opid that fails after submission — still surfaces via
the existing send-progress notification; only the submit gate is reflected here.)
Also removes the now-unused chat_readonly_note string (the composer replaced the
read-only footer). Test: delivery status round-trips through the DB. Verified:
Linux + Windows build with chat ON, ctest 100%, hygiene clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The full-node daemon unconditionally shuffles a transaction's shielded outputs
(transaction_builder.cpp ShuffleOutputs, to hide the change position), so the
two 0-value HushChat memo outputs land at random note positions — the header is
NOT guaranteed to precede its payload. The receive parser paired a header only
with the NEXT payload in position order, so a shuffled send failed to pair (the
message was silently lost) ~50%+ of the time.
groupHushChatMemoOutputs now holds an orphan payload (one seen before its header)
and pairs it with the header when it arrives, so header↔payload pair regardless
of on-chain order. A chat tx carries exactly one header + one payload (the harvest
already skips change outputs), so this is unambiguous.
This is the complete fix for the ObsidianDragon ecosystem: full-node and lite both
receive through this one C++ parser, so full-node→full-node and full-node→lite now
work despite the shuffle, with no daemon fork and no lite-backend change. Context:
SilentDragonXLite (which hardcodes payload position==1 and forced the daemon/backend
question) is being replaced by ObsidianDragonLite. Lite send never shuffles.
Tests: shuffled receive (payload-before-header, and change-interspersed) pairs +
decrypts; all existing ordered-input tests unchanged. Verified: Linux + Windows
build, ctest 100%, hygiene clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Make composed messages actually reach the wire: broadcastChatMemos sends the
header + payload as two 0-value memo outputs to the peer's z-address (header
first, the lower memo position), on both variants.
- chat_outgoing: chatSendOutputs(memos, utf8Prefix) — the pure, testable memo
encoder. Full-node memos get a "utf8:" prefix (dragonxd rejects raw JSON and
then UTF-8-encodes on-chain, byte-identical to SilentDragonXLite's
Memo::from_str); the payload is NOT double-hex. Lite memos are raw UTF-8 (the
backend does Memo::from_str directly). Header is always output 0.
- App::broadcastChatMemos: full-node builds a two-recipient z_sendmany array
(amount 0, from the spendable reply z-address, default fee) via submitZSendMany
with markFeeGapRetry=true — deliberately, to suppress the fee-gap auto-retry,
which rebuilds a single-recipient tx and would drop the payload output. Lite
routes to broadcastChatMemosLite (two 0-value LiteSendRecipients, raw memos;
the backend accepts duplicate addresses + 0-value for exactly this pattern).
- Encoding + design established by a four-codebase mapping (wallet, daemon, SDXL,
lite backend) and an adversarial review of the wiring (0 confirmed findings).
- Tests: chatSendOutputs (utf8:-prefix + header-first for full-node, raw for
lite) + on-chain round-trip (strip "utf8:" -> the harvest parser re-pairs and
decrypts).
Remaining as a LIVE test (cannot be proven from source): that dragonxd returns
the memo under memoStr verbatim, that recipient-array order maps to note
position, that a 0-value memo-only tx relays, and full SDXL<->DragonX interop.
Gated by DRAGONX_ENABLE_CHAT (default OFF). Verified: Linux + Windows(mingw)
build with chat ON, ctest 100%, hygiene clean; caches restored to the OFF default.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add the outgoing side of HushChat: compose messages and start conversations.
The wire construction is the byte-exact inverse of the receive parser and is
proven by a self-consistent round-trip (build → parse → decrypt).
- src/chat/chat_outgoing.{h,cpp} (pure): buildOutgoingMessage encrypts via
encryptOutgoing and buildOutgoingContactRequest carries plaintext; both emit
the header memo JSON ({h,v,z,cid,t,e,p}, which nlohmann serializes
alphabetically to match SilentDragonXLite) + the payload memo, validating the
512-byte memo limit, the 64-hex peer key, and the "no leading '{'" rule for
request text. My public key goes in header "p"; the peer's key is the
encryption recipient.
- ChatService: composeMessage/composeContactRequest (encrypt with the held
identity) + recordOutgoing (echo an Outgoing ChatMessage into the store + DB —
we never harvest our own sends, so this is the only local record).
- App: sendChatMessage(cid,text) sources the peer z-addr + public key from the
conversation, composes, and records a random-id echo; startChatConversation
(zaddr,text) mints a random cid + composes a plaintext contact request;
chatReplyZaddr() picks a spendable z-addr. broadcastChatMemos() is the Phase-5
transport seam (network delivery + real-SDXL interop verification land there).
- Chat tab: a message composer (shown once the peer's key is known, else a
"waiting for reply" hint) + a "New conversation" modal that sends a contact
request to a z-address.
- Tests: outgoing round-trip through the receive parser + decrypt, contact-request
passthrough, validation guards, and ChatService compose + recordOutgoing echo.
Adversarially reviewed (crypto/interop, secret hygiene, logic, UI) — 0 findings.
Gated by DRAGONX_ENABLE_CHAT (default OFF). Verified: Linux + Windows(mingw)
build with chat ON, ctest 100%, hygiene clean; caches restored to the OFF default.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Swap the in-memory-only chat store for durable sqlite persistence with real
per-transaction timestamps, encrypted at rest under a key derived from the
wallet's own seed (no passphrase, works on encrypted and unencrypted wallets).
- ChatDatabase (src/chat/chat_database.{h,cpp}): sqlite store mirroring
data::TransactionHistoryCache. unlockWithSecret(seed) derives a 32-byte AEAD
storage key and a wallet-partition tag via domain-separated keyed BLAKE2b
(generichash) contexts. Every record — bodies, peer z-addrs, threading,
timestamps — is crypto_aead_xchacha20poly1305_ietf-encrypted with a random
nonce and the wallet tag as associated data; even the dedup key is a keyed
hash of txid+position, so nothing about your conversations is in cleartext on
disk. Rows are partitioned per-wallet; a different seed sees nothing. Messages
are decrypted once at ingest then re-encrypted under the storage key, so
load() needs only the storage key, not the chat identity.
- ChatService: ingest() now stamps each message with its own transaction time
(txid->time map + fallback) and writes new (store-deduped) messages through to
the database; loadFromDatabase() rehydrates the in-memory read model on unlock.
- App: unlock the chat DB with the same seed in provisionChatIdentityFromSecret
and load prior history; lock the DB + clear the decrypted in-memory store on
relock and on lite-controller rebuild.
Adversarial review (4 confirmed findings, all fixed): don't provision if the
wallet locks mid-fetch (re-check isLocked at completion); wipe the serialized
plaintext temporary in append(); trim the seed into a separate fully-wiped
buffer (no residue past a shrunk size()); scrub the mnemonic copy in the RPC
json result.
Tests: ChatDatabase round-trip (persist/reload, field + order fidelity), dedup,
per-wallet isolation, lock inertness, and ChatService write-through + reload
without an identity. Gated by DRAGONX_ENABLE_CHAT (default OFF). Verified:
Linux + Windows(mingw) build with chat ON, ctest 100%, hygiene clean; caches
restored to the OFF default.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Phase 1 Steps 4-6 (the receive pipeline, minus the App/sync wiring which
lands next).
- chat_message: the in-memory ChatMessage model (direction, kind, txid, cid,
peer zaddr/pubkey, body, timestamp, payload_position). No libsodium.
- chat_store: threads messages by conversation_id and dedups by
(txid, payload_position) so re-scanning the chain never double-inserts.
- chat_service: owns the long-lived chat identity keypair (move-disabled,
wiped on destruction/clear) and the store. ingest() decrypts each Message
(drops undecryptable ones silently — no plaintext/memo logging), passes a
ContactRequest's plaintext through, and threads the result. No-op without an
identity.
Tests: a metadata batch decrypts into a threaded conversation; re-ingest
dedups; a contact request carries through; a wrong identity decrypts nothing;
no-identity ingest is a no-op.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Phase 1 Step 1 (the linchpin). The receive path had nothing to decrypt:
HushChatTransactionMetadata carried only payload_position/size, and the
extractor discarded the paired payload memo + the header's "e"/"p". Widen the
metadata with sender_public_key_hex ("p"), secretstream_header_hex ("e"), and
payload_memo (ciphertext hex for a Message; request text for a ContactRequest),
populated from the already-parsed header/payload pair.
Add an end-to-end receive-path test: derive two identities, encrypt Alice->Bob,
wrap it as a HushChat header+payload memo pair, run it through
extractHushChatTransactionMetadata, and decrypt straight from the carried
metadata. Also asserts the harvest yields nothing when the feature is disabled.
Fields are only populated when featureEnabled (the extractor already returns
early otherwise), so an OFF build produces no decrypt material.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Phase 1 crypto foundation (gated by DRAGONX_ENABLE_CHAT; the pure primitives
are always compiled + unit-tested, the feature gate lives at the future
service layer).
- chat_identity: derive the DragonX-native X25519 (crypto_kx) identity from a
stable per-wallet secret via a domain-separated keyed BLAKE2b KDF
(crypto_generichash, context "DragonX-HushChat-Identity-v1") into a clean
32-byte crypto_kx seed. Deterministic; skips SDXL's UTF-8-hex-seed quirk
(that's the Phase-4 import path). Per §5.6.
- chat_crypto: encryptOutgoing (server_tx) / decryptIncoming (client_rx) via
crypto_secretstream_xchacha20poly1305, byte-exact per Appendix A.3/A.4 so
DragonX interoperates with SilentDragonXLite. Single chunk, TAG_FINAL;
decrypt enforces both the Poly1305 tag and TAG_FINAL (stricter than SDXL).
Every session key / seed / stream state / plaintext scratch is sodium_memzero'd
on all paths; no secret is ever logged.
- Tests: encrypt->decrypt round-trip (incl. empty + UTF-8), identity
determinism, feature-gate + empty-secret handling, and malformed/tampered/
wrong-key inputs all fail safely.
Adversarially security-reviewed (3 lenses: secret hygiene, crypto correctness,
SDXL wire-interop) — confirmed byte-compatible with the SDXL reference in both
directions. Fixes from review: check the secretstream_push return before
reporting Ok; allow the empty-plaintext ciphertext (== ABYTES) so encrypt/decrypt
round-trip symmetrically; document the caller-owns-and-wipes secret contract.
Interop caveat: round-trip proves self-consistency; a real captured SDXL
message is still needed to prove wire-compat end to end.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two dead-code removals surfaced by the audit; no runtime behavior change.
1. Delete the lite-lifecycle "readiness" scaffold (~1,586 lines). The pure
dry-run evaluateLiteWalletServerLifecycleReadiness (whose own status text
says "wallet lifecycle execution is still disabled in this scaffold") had
zero callers; executeLiteWalletServerSelectionUi had zero callers; and
executeLiteWalletLifecycleUiRequest's only caller was a fallback in
settings_page reached only when app->liteWallet() is null. Deleted
lite_wallet_server_lifecycle_readiness.{h,cpp} and the readiness /
UI-execution machinery + enums + translation tables from both adapters,
keeping the genuinely-live helpers (liteConnectionSettingsFromAppSettings,
applyLiteConnectionSettingsToAppSettings, redactLiteServerSelectionValue,
and the LiteWalletLifecycleUiExecutionInput DTO the live create/open/
restore path still uses). The null-backend fallback now sets a plain
"Lite wallet backend unavailable" status. This is exactly the
readiness/scaffold pattern CLAUDE.md forbids regrowing.
2. Split the HushChat fixture/capture-manifest/seed-projection tooling
(~2,050 lines, incl. the libsodium seed-projection) out of
chat_protocol.cpp (1854 -> 284) and chat_protocol.h (586 -> 105) into a
new chat_fixture_tooling.{h,cpp} compiled ONLY into the HushChatFixtureCheck
dev tool — never into the app, lite, or test binaries (verified via nm).
The app keeps only the runtime slice (memo parsing + tx metadata
extraction) that network_refresh_service actually calls. Also moved the
decrypt-preflight/hex helpers, which likewise had no runtime caller.
Note: the HushChat split is on gated-off experimental code and should be
coordinated with the pending dormant-HushChat-content handling (commit
af06b8b) before the lite PR — done here as a pure mechanical split, no redesign.
Full-node + Lite + HushChatFixtureCheck build clean; ctest 1/1; hygiene clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Preserve the previously-uncommitted lite wallet implementation and related dev WIP
under version control:
- src/wallet/ lite services: client bridge, bridge runtime, connection, lifecycle,
sync, gateway, result parsers, state mapper, artifact contract/resolver, refresh
services, UI adapters, wallet_backend/capabilities. (Includes two small M1 fixes:
lifecycle walletReady now parses the response; default chain name -> "main".)
- src/chat/ chat protocol; tests/fixtures/ (lite + hushchat); tools/hushchat_fixture_check.cpp;
scripts/build-lite-backend-artifact.sh.
- Pre-existing modified app_network/security/wizard, network_refresh_service, sidebar,
mining_tab, bootstrap dialog, and version headers captured as-is.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
