Skip to content

feat(chat): restore separate Chat authority integration above PolkaVM - #255

Open
replghost wants to merge 132 commits into
feat/pvm-wasmfrom
feat/chat-v2-host-runtime
Open

replghost wants to merge 132 commits into
feat/pvm-wasmfrom
feat/chat-v2-host-runtime

Conversation

@replghost

@replghost replghost commented Sep 17, 2026 •

Copy link
Copy Markdown
Collaborator

Current repair qualification — 2026-10-02

  • Frontend head: 6ef2aa616b6e771c236a37f36743212472ad0b38. Prior main baseline: f4012c50c3400d1186c332ad2ae50298cefd153d.
  • Native source and canonical artifact-generation pin: host-rust-core #709 at 726ef76bcce42b91d664d753ac4225df0a093472 (chat). Complete matching SDK/host packages remain 0.23.0; no mixed or hand-patched Wasm bundle. Generated client bytes are unchanged. Production: wasm-signing-host; separate testing bundle: wasm-signing-host,test-host.
Canonical artifact SHA-256
Client npm archive d8e759c51c8e49dd829efbd3d5227217aeac9bd8d9b8952397716fe8bdd8f95c
Client dist/generated/client.js 6c48f48837ceda0c80163dbbc208a365e44230fd55e4b472b634bb3ac2a77aa3
Host npm archive 8ecea2f14eecf6bdde983f04c1d28d1c7c6bffddef639d6a31cf5a7f38ec50b9
Browser signing Wasm 0f10b028409e4d17f6256f3f4fc08fec1fd9bf6a2102f19e77e2e91a567e4a6f
Local QA CLI e21b93f44979f8a46f5f066e1dca6259984334ff4f41dd1e71f09781b5402efa

Native correction and causal limits

Bulletin mortal signatures now anchor to a finalized checkpoint, while nonce/runtime state comes from the freshest available signing snapshot. A checkpoint at least 64 blocks behind is rejected before broadcast. Regression coverage verifies the actual signature with a newer nonce and older finalized checkpoint, and the expiry boundary. Existing transaction retries/deadlines are unchanged; no chain error is suppressed. The affected test fixture uses parking_lot::Mutex.

The former Seity Extrinsic marked as invalid run used noncanonical best-block anchors, but retained evidence lacks the signed bytes and typed pool reason. Fork-sensitive mortality was a real correctness hazard; it is not claimed as the conclusively proved cause of that historical failure. Prior hosted Factory/Genesis timeout causes also remain unproved. No broker/readiness workaround, prewarming, extra retry, or timeout increase was added.

Final exact-head hosted checks

Workflow Result
Secret Scan success
Dependency Audit success
Bundle Size success
Static Analysis success
Performance Tests success
Tests success
  • Functional: 89 passed, 3 skipped.
  • E2E: 42 passed, 20 skipped.
  • Performance: 3 passed in the cold-start suite. This hosted suite measures cold start, not the Doom backend matrix.
  • No flaky-test summary in this exact-head hosted Tests log.
  • Corresponding native CI: success. Each of five native variants passed 21 Bulletin regressions and all-target Clippy; base full library passed 1,246 tests.

Real-runtime local verification

All six consumers passed configured build, typecheck and lint. The later README-only formatting correction passed targeted Prettier checks and did not change runtime source or canonical native package bytes. Original full Wallet→Chat→Seity test sequences, without test retries, added waits, or prewarming:

Local original sequence Tested frontend head Result, retries=0 Genesis after navigation Factory Submit
#238 0a24db4aa189adad7b93f9ff508c7949e39196ee 42 passed, 20 skipped; no failed/flaky cases 1107ms 9625ms 6057ms
#255 301a235ba13248ef4281e9e3d859919fb6cfbe06 42 passed, 20 skipped; no failed/flaky cases 1045ms 11023ms 13548ms
#287 d3dc8a41807c36b8bcdf001c47eccb9735305499 42 passed, 20 skipped; no failed/flaky cases 1029ms 6496ms 5530ms

Wallet's tested head is final. Chat and Seity tested heads precede documentation-only formatting merges; native/Wasm pins and runtime code are unchanged. Factory, Genesis-after-navigation and Submit passed on all three. These bounded passes do not establish the cause or elimination of the older hosted timeouts. Prior SharedWorker lifecycle, Chat/profile, Peer and JAM runtime proofs remain historical at their recorded heads; they were not all rerun on these artifacts.

Final hosted retest: failures remain

Frontend PR Tests workflow Functional E2E
#238 success 79 passed 42 passed, 20 skipped
#185 success 89 passed, 3 skipped 41 passed, 1 flaky, 20 skipped
#255 success 89 passed, 3 skipped 42 passed, 20 skipped
#287 failure 85 passed, 4 failed, 3 skipped 42 passed, 20 skipped
#290 success 87 passed, 2 failed, 3 skipped 42 passed, 20 skipped
#291 failure 86 passed, 5 failed, 3 skipped 41 passed, 1 flaky, 20 skipped
  • PVM and JAM Genesis Hash still flaked after navigation: the product stayed pending for the helper's existing 20 seconds; the existing CI retry passed. The helper reports this timeout as error; it is not evidence of a chain rejection. Retained traces include People-chain/local development sockets but do not expose the Asset Hub gateway WebSocket exchange, so these new failures do not establish an upstream or broker cause.
  • Peer: two navigation product-render timeouts at 45 seconds. The existing threshold tolerates these, so its green workflow is not a clean functional suite.
  • Seity: cache-off RPC-gateway resolution timed out; two navigation renders timed out; another navigation case reported failure to connect to trusted provider paseo-bulletin-next-ipfs.polkadot.io. The zero-failure host-settings gate correctly failed.
  • JAM: cache-on reload and two navigation cases reported that same provider connection error; two more navigation cases timed out. The zero-failure host-settings gate correctly failed. These logs do not establish whether the underlying cause was provider availability, runner networking, CORS, or another transport failure.
  • All six final Static Analysis and hosted cold-start Performance workflows passed; all five native CI workflows passed. That does not clear the red functional gates, Genesis flakes, or Doom matrix. No additional CI rerun, timeout/retry increase, threshold relaxation, or speculative transport patch was made.

Doom: criterion corrected, backend matrix still unqualified

The user explicitly approved 35 FPS sustained over 30 seconds, with one frame of sampling-boundary tolerance: frames + 1 >= elapsedMs * 35 / 1000. This replaces the instantaneous FPS >= 35 sample; it is an acceptance-criterion change, not a runtime speedup. Runtime and displayed FPS are unchanged; raw samples are not rounded to force a pass. Update p95 <28.6ms, cold/warm first-frame limits <3,000/<1,000ms, audio and translation-cache checks remain. The revised official RPC benchmark passed once before the vendor replacement.

The later isolated matrix used the rebuilt base pair at frontend 238ab5fa739788b2eb948ba1f807a1876edbacad, with fresh owned Chrome profiles and no concurrent builds/E2E:

Backend Observed result Qualification
smoldot-direct 1,050 frames / 30,001.1ms; p95 17.7ms; cold/warm 202.9/152.8ms; audio/cache passed PASS under the approved criterion
smoldot-shared-worker 1,038 frames / 30,001.4ms = 34.5983854 FPS; p95 17ms; cold/warm 186.4/158.4ms; other checks passed FAIL: cadence deficit retained
rpc-gateway Warm-readiness helper timed out. Its null snapshot conflated missing frame, missing canvas and evaluation errors, and earlier cold/cadence values were not retained by that helper NOT QUALIFIED; iframe disappearance is not established

Distinct diagnostics found no presentation loss in a later instrumented 24-second shared-worker sample and observed a successful hidden RPC new-document reload. Neither diagnostic supersedes the failures or qualifies performance. No evidence-proved runtime fix, guest rebuild, speculative tuning, or further unchanged gate rerun was made. The original strict RPC failure and every new failure remain retained. Next causal capture must instrument the original early 30-second window and distinguish warm lifecycle states before proposing a runtime fix.

Retention and rollout boundary

Local verification used paseo-next-v2, the pinned host-playground fixture f56294cea4430163bf16ec068844b1327441073c, and existing private QA identities. All three corrected sequences paired on their first existing setup attempt. A prior local launch failure and a misconfigured Previewnet-product run (34 passed, 20 skipped, 8 Chain/Contract failures) are retained separately, not presented as reproductions of the hosted Genesis timeout. Private traces/auth/signers were not published.

The first repair heads also exposed a README formatting failure (corrected by documentation-only commits) and a PVM cache-unit-test 5,000ms timeout. No cache runtime/test change, limit increase, or causal claim was made for that isolated timeout. Older in-flight runs superseded by the formatting correction remain recorded as cancelled, not passed.

No PR merge, force push, deployment, environment approval, SDK/npm/product/guest publication, or rollback was performed by this repair. JAM remains JAM-TEST-INSTANCE, never JAM-PUBLIC-DEVNET. The user-owned deploy: paseo.fyi label is retained; repair-triggered deployment runs 36970790399 and 36971286469 were cancelled before deployment. The earlier rollout audit is retained below: an older workflow deployed 56cea5d37577bf82684fb5c6b6e4ba81d2142938; this record does not claim live remained unchanged historically.

Historical integration qualification (superseded; retained verbatim)

Current merged-main refresh and qualification — 2026-10-01

Canonical artifact SHA-256
Client npm archive d8e759c51c8e49dd829efbd3d5227217aeac9bd8d9b8952397716fe8bdd8f95c
Client dist/generated/client.js (not dist/index.js) 6c48f48837ceda0c80163dbbc208a365e44230fd55e4b472b634bb3ac2a77aa3
Host npm archive a725bcbd6ff17a92bea89bb11868a45b6b3fdb178cb037cb00da1c46105526c9
Browser signing Wasm 0778dfa5255092bbfc7607dc226d1fc9a2d2d0315ffb669258341fa307556485

Stack and hosted qualification

Wallet → PVM → Chat → Seity; PVM → Peer; JAM contains both Seity and Peer. All updates used fresh-fetch ancestry checks and ordinary pushes. User worktrees, metadata and canonical repositories were preserved.

  • #238: 70845db96bdd21409310daeff5e471b74ab5a1c9.
  • #185: 351e5d73fab803a0fd2a4007ce2d2557aebaff05.
  • #255: bb9ad3d8921c63cf62c98b9029289f458b1a67a2.
  • #287: b03e83616857f4e4745d479776dde43cdfa6267a.
  • #290: 44ed4f830a2c8f86990e02fb061c802aa389b0bc.
  • #291: dcdef40499119183aed82a39f784ef740d494341.

Exact-head hosted results: functional 89 passed, 3 skipped; E2E 41 passed, 1 flaky, 20 skipped; 3 passed in the cold-start suite. Tests run.

Hosted E2E retained one flaky Chain → Chain Spec: Genesis Hash case: it failed on the first attempt and passed the existing CI retry. No retry setting was changed.

Performance Tests runs the cold-start suite, not the strict Doom matrix below. Existing thresholds, retries, deadlines and skips were not relaxed. Native current-head Core CI has 22 passed jobs and 2 path-filter skips; full iOS application CI is not claimed for the latest native heads.

Local and actual-runtime evidence

  • Typecheck, lint, format, full unit suites and builds passed across all six frontend variants. The final runtime-code cutover is unchanged; later test/docs-only fixes removed the obsolete whole-resolver mock at PVM and moved fresh bitswap module loading into per-case setup before provider fixtures are installed. The timed-out cold-import continuation can no longer issue requests through the following case’s provider. All 17 targeted bitswap cases and the root full gates passed without changing a test timeout, retry/discovery bound, assertion value or skip. The actual stopped-follow browser regression passed on every corrected descendant with the unchanged 10s deadline.
  • Wallet functional suite: 79/79 passed. Corrected PVM full functional suite: 89 passed, 3 existing artifact-dependent skips. Combined JAM functional suite: 91 passed, 3 existing artifact-dependent skips, including both strict cached-host upgrade scenarios; final JAM changes after that full run are ancestry/docs-only. Combined local E2E: 42 passed, 20 existing skips. These are separately scoped runs, not fabricated per-branch full-suite counts.
  • Real RPC fault injection stops a configured node’s active chainHead storage read; the application loads with exactly one fresh follow and no domain error. The canonical resolver owns the one retry and the remaining original sync budget. An earlier hosted Peer run failed before collection because the obsolete mock duplicated this regression title; that integration error was fixed at PVM and propagated through all descendants, without changing a gate.
  • Browser/native terminal recovery: the original CoreWorker/client/transport and native connection remained open after a real uncaught SharedWorker error. The old follow stopped; a later real chain query recovered through a new URL generation in 8,147ms, and a new follow on the same client delivered 76 live NewBlock events. Same-partition cross-tab sharing was verified before and after replacement. This does not claim continuity of the stopped follow. Evidence: main-refresh-terminal-verified-pvm.json / screenshot.
  • Real transient RPC closure/replay was also observed with one consumer receiving post-reconnect statement data and remaining alive for 130s. Evidence: main-refresh-pvm-transport.json / screenshot.
  • Feature-specific proof: Actual canonical SDK Chat Initialize passed over the authenticated product port. Separate ChatIdentityAuthority Allow-once consent was required. After a full host reload and new Allow-once consent, public device metadata was identical. Evidence: main-refresh-native-chat.json. No published guest/product was created.
  • Live probes used the canonical feature bundles before final test/docs-only changes; temporary drivers/profiles were removed and public JSON/screenshots retained beside the refresh worktrees. Testnet-only identities, statements, preimages and allowances were permitted; no mainnet write was performed.

Strict Doom matrix — retained failure

Fresh isolated visible headless Chromium profiles; cold cache miss, real compiler, nonzero audio, and 360 frame updates for each backend. Warm reload uses a new document in the owned iframe: cache hit and zero translation. Gates remain FPS ≥35, p95 <28.6ms, cold first frame <3000ms, warm first frame <1000ms.

Backend FPS p95 ms Cold first ms Warm first ms Result
rpc-gateway 34.54231433545555 13.1 163.8 128.6 FAIL: FPS
smoldot-direct 35.06721215581914 13.2 173.7 110.1 PASS
smoldot-shared-worker 35.48895899032775 11.8 164.3 100.3 PASS

Evidence: main-refresh-isolated-doom-{rpc-gateway,smoldot-direct,smoldot-shared-worker}.json and screenshots. Earlier hidden-tab and concurrent-QA failures remain recorded; no clock/rounding/threshold change or failure-only rerun was used to erase them.

Other retained limits

  • Latest local Seity E2E attempt: 41 passed, 20 existing skips, 1 Submit failure (Extrinsic marked as invalid). One UI call caused the existing two native broadcast attempts; both validated then invalidated. No matching included Submit was found in the inspected canonical block range. [INFERENCE] Noncanonical mortality anchors may explain invalidation, but the exact cause is unproved without the unretained signed extrinsic bytes. Evidence: main-refresh-seity-submit-diagnosis.json. Passing hosted E2E does not erase this local failure.
  • Earlier Seity Factory hit its unchanged 30s product deadline; its original payload was later included after 47,516ms. A separate unchanged manual Factory attempt passed in 10.918s, and Factory passed in the later full local suite. Evidence: main-refresh-seity-factory-diagnosis.json. No added retries, deadline changes, or unsupported cause claim.
  • Seity’s positive encrypted-content proof is not a full registered-user profile/contact-sharing round trip. Peer/JAM proof used the installed PVM host-frame-request path with a throwaway canonical SDK driver, not a newly published guest. Some validators refused connections; successful connections used actual certificate-verified WebTransport.

Rollout boundary

Not an all-green rollout qualification. The refreshed heads were not deployed, and this qualification granted no environment approval. No PR merge, SDK/npm publication, or guest/product publication was performed. JAM remains JAM-TEST-INSTANCE. All three refresh-triggered JAM Deploy runs were cancelled before rollout; latest final-head cancellation. Other branches have no deployment labels.

The live environment changed during qualification: older Deploy run 36945647296, attempt 3 was approved under GitHub account replghost, explicitly checked out baseline 56cea5d37577bf82684fb5c6b6e4ba81d2142938, and recorded deployment success at 2026-10-02 01:19:01 UTC. Its later published-product smoke failed. The operator/client/session behind that account is unproven; this qualification granted no approval and performed no rollback. The observed live content hash changed from 4a7caf047fb7f350c1833039b93ba634698a47c56f051b8c7fc0525d4c59a5c8 to 5882695ee20df5d24a6ed2feb9e997f176fe27ecf4eafe742fabb39eba92885f; these are content hashes, not Git SHAs, and their exact byte-level mapping to a source commit is unproven.

Earlier deployment authorizations/results below are historical evidence, not approval to deploy this refreshed stack.

Historical integration qualification (superseded; retained verbatim)

Current-main integration qualification (2026-10-01)

  • Published head: 19253c28214900ce1a41a727e32f29672d0f90fb. Includes main 74bb603e518ace149a846124f6e7353aca198c57, propagated through feat(wallet): add an opt-in debug-only browser test wallet #238 → feat(polkavm): run App Manifest v2 runtimes in Dotli #185 → feat(chat): restore separate Chat authority integration above PolkaVM #255 → feat(profile): show product-referenced Seity profiles in a host drawer #287 and feat(polkavm): run App Manifest v2 runtimes in Dotli #185 → feat(polkavm): JamPeerTransport over WebTransport for JAM apps (layer atop #185) #290, then combined in chore(deploy): Seity layer + JamPeerTransport (#287 + #290) #291. All six heads are published by normal ancestry-guarded pushes; no history rewrite.
  • Preserves main’s Astro/Solid architecture, explicit package APIs, strict NodeNext, one shared native page core, execution-local callbacks/consent, and host-owned block/archive cache. Retiring an execution does not retire core-owned authentication. Concurrent quiet-RPC/transport-loss fixes for Trusted Providers silently loses Statement Store subscriptions after 120 seconds of inactivity #308 are preserved throughout the stack, including Chat/Seity/JAM HOP support.
  • Canonical client/host SDK 0.23.0 source: 437a46c5af88b3c4a962d763f8fa5732e9c48183; complete feature-specific artifacts, signing-only browser Wasm, and separate testing Wasm. Native base #540 CI and Peer #1010 CI are green. No npm or guest publication.
  • Native integration proof (before the stylesheet-only follow-up): All 44 workspace tasks passed uncached after the concurrent RPC fix, followed by root checks/format and a fresh debug build. Actual egui-chat cancel → redraw → Retry → new sign-in → cancel, including mobile layout, passed on RPC and SharedWorker (4/4, zero browser exceptions). Subsequent follow-ups only adjust browser/test synchronization and remove a duplicated incidental-default case, with maintenance guidance.
  • Final-head CI gates pass, including Unit, Functional, and E2E. Functional: 89 passed, 0 failed, 3 skipped. E2E: 42 passed, 20 skipped. The three functional skips are existing artifact-dependent cases. No failure threshold or retry policy was relaxed.
  • Limits: Signed-out Chat proof does not claim a payment, identity registration, or messaging write.
  • Fixture-only CI follow-up: persisted settings can precede address-bar canonicalization. The two browser assertions now await the actual URL condition; no timeout increase, production change, or threshold relaxation. Ten Wallet and ten combined production browser iterations passed, followed by host type/lint and repository formatting checks. The corrected settings cases passed final-head hosted qualification.
  • Debug stylesheet follow-up: actual deployed inspection reproduced a 156 px unused gap caused by measuring the lazy debug panel before its CSS arrived. The Wallet-root fix imports CSS with the lazy module before mounting; obsolete manual-link injection and test workarounds are removed. The delayed-CSS behavioral regression failed before the fix and passes locally and on every final hosted head. All 91 focused debug-panel tests passed. Type/lint/format checks passed. Actual local RPC and SharedWorker Doom now meet the 360 px panel without a gap. Native artifacts are unchanged by this follow-up.
  • Rollout: Deployed only to paseo.fyi at combined chore(deploy): Seity layer + JamPeerTransport (#287 + #290) #291 117ef1eb2928a470e65e48c05b5bf22e8ab29c02 (successful deployment and smoke, deployment ID 6778272827). Hosted smoke passed 8/8 published products and 19/19 wallet-free capabilities. Independent fresh HTTPS Chromium sessions confirmed Build 0.6.0 (117ef1e) and playable Doom with advancing frames/input/audio on smoldot-direct, RPC gateway, and an actual SharedWorker, with zero browser exceptions. All three screenshots and geometry checks show the 384 px frame meeting the 360 px debug panel without a gap. An actual retained production c4a08e9 session waited for consent, navigated exactly once after Reload, retained the URL plus localStorage/IndexedDB/product-cache markers, and resumed playable Doom on 117ef1e with an activated controlling worker, no waiting worker, and zero exceptions. This is an actual release upgrade, not a substituted worker-script test. Local JSON/screenshots are retained under dotli-community-worktrees/paseo-fixed-doom-* and paseo-debug-fix-retained-*. No other environment was approved; no PR merge, force push, npm release, or on-chain guest publication.

Earlier sections are historical implementation and qualification records.

Latest dependency refresh qualification (2026-09-30)

  • Published head: a0de426a7923506bd32a6f6c1f5670527cfefd53; normal merges and ancestry-guarded pushes only, no history rewrite.
  • SDK: TrUAPI client + host 0.23.0, source 437a46c5af88b3c4a962d763f8fa5732e9c48183. Feature-specific 0.23.0 client/host/WASM artifacts remain unchanged; this source pin is intentionally not relabeled as the refreshed upstream branch head.
  • Provider: 0.3.1, including buffered-RPC lifecycle handling. Previewnet's four genesis hashes now match the provider catalogue and upstream reset #995; DotNS addresses/slots, suffix, and endpoints are unchanged.
  • Real-browser Previewnet check: relay, Asset Hub, Bulletin, and People each returned its configured genesis and synchronized health with a peer. This is provider-level proof, not Previewnet product-publication or identity-registration qualification.
  • Local migration format/lint/typecheck/unit/build checks passed. The final network change additionally passed all 54 configuration tests, config lint/typecheck, and the four-chain browser smoke.
  • Local lint, typecheck, unit tests, and build passed; the Chat-specific SDK artifacts were preserved.
  • Final hosted checks passed for this head, including E2E: https://github.com/paritytech/dotli-community/actions/runs/36750583417. Functional results: 47 passed, 0 failed, 3 skipped. The E2E workflow installs rustfmt for the checked-out signing-host source's declared nightly-toolchain (older feature sources retain their existing nightly behavior).

Earlier implementation and qualification history follows; the versions and heads above describe this refresh.

Summary

Chat authority integration above PolkaVM #185, paired with host-rust-core#709. This is the active Chat layer replacing historical #236.

SDK provenance

  • @parity/truapi and @parity/truapi-host: 0.23.0.
  • Matching client, host, and wallet WASM source: 437a46c5af88b3c4a962d763f8fa5732e9c48183 in host-rust-core#709.
  • Exact archives, generated-client/WASM digests, build features, and license records are in vendor/truapi-host.lock.json. Wallet WASM enables wasm-signing-host, not test-host.

Artifact boundary

The final SDK refresh pins host-rust-core#709 at 437a46c5af88b3c4a962d763f8fa5732e9c48183, including the concurrent Contacts integration and subsequent native Chat review fixes. Client and host archives came from the canonical generated client, TypeScript builds, and matching wallet/testing WASM outputs of that exact source. The generated client digest remains unchanged; the host archive and wallet WASM changed. Their archive and artifact digests are recorded in the lock.

The user explicitly selected deployment of the newer sharing APIs on experimental paseo.fyi. The combined #291 deployment is 76eb846e06713679891c6e0cc17a65919909eced, built from host source af087f734be4d65d9363e68e43bf134145ae0085; its evidence is separate from this branch-only CI result.

Experimental execution gate

PolkaVM execution defaults on for test environments (paseo.fyi, paseo.li, previews, and localhost) and off for production dot.li. Settings → Experimental → PolkaVM apps explicitly overrides either default; existing saved opt-outs and opt-ins remain authoritative. This is an execution gate, not build-time exclusion of runtime assets or Firebase-controlled eligibility. Product permissions remain separate from the execution setting.

Qualification

Branch-CI-qualified head: cec01502c69856e3599c65b8af4d6da46e9ea06d.

Current-head checks pass, including unit, functional, and live E2E, type-checking, production builds, and full-history secret scanning. This qualifies branch CI, not a deployment or a fresh-browser smoke at this head. Deployment evidence is qualified separately in #291.

After pulling these local-file SDK dependencies, use bun install --force --frozen-lockfile to replace Bun's cached packages.

Earlier local Chromium verification rendered the real published PVM guest and completed a canonical SDK handshake through its live host port. The current paseo.fyi deployment is the combined #291 layer, not this branch alone.

Chat 0.2.1 publication transactions and prior exact-source native evidence remain recorded in #255 (comment). No wallet creation, funding, payment, or on-device Chat message was performed for this refresh.

Startup applies URL-selected protocol settings before wallet/debug hooks and drains shared-mode writes before iframe replacement or startup-triggered reload. Testnet published-product smoke starts without opting in; production dot.li remains default-off and its smoke uses Settings and Save & Apply, and all existing rendering/input/audio/GPU checks remain intact. Failure-only deployed smoke traces and screenshots are retained for three days.

Secret scanning retains the default rules and full-history scan. The public resolver mapping-slot vector is exempted only when both its exact value and test path match; a boundary smoke confirms changed values and the same value elsewhere remain detectable.

The current combined default-on release, 76eb846e, passes 8/8 published products without a testnet opt-in and 19/19 wallet-free capabilities. Independent fresh Chromium observed that exact release, ran Doom with the default enabled, and preserved an existing testnet opt-out across reload. This is combined #291 deployment evidence, not a deployed smoke of this branch alone.

@replghost
replghost requested a review from a team as a code owner September 17, 2026 02:23
@replghost replghost added the deploy: paseo.fyi Deploy this pull request to paseo.fyi label Sep 17, 2026
@github-actions

github-actions Bot commented Sep 17, 2026 •

Copy link
Copy Markdown
Contributor

Bundle Size Report

Warning

Bundle budget exceeded — not a blocker, but please confirm the regression is intentional.

  • host-eager-path exceeded budget: 170702 B gzip > 168000 B
  • sandbox-eager-path exceeded budget: 105644 B gzip > 87000 B

Eager path (module entry + modulepreloaded chunks):

Path Raw Brotli Gzip
host/(eager path) 434.5 KB (+37.5 KB) 146.0 KB (+14.2 KB) 166.7 KB (+16.2 KB)
sandbox/(eager path) 316.8 KB (+81.9 KB) 89.1 KB (+22.1 KB) 103.2 KB (+25.6 KB)

Chunks over 500 KB:

File Raw Brotli Gzip
host/assets/bridge.js 793.8 KB (+696.8 KB) 173.1 KB (+151.6 KB) 222.8 KB (+196.7 KB)
host/assets/truapi_provider_bg.wasm 4.01 MB 955.7 KB 1.27 MB
host/assets/truapi_server_bg.wasm 4.32 MB (+1.63 MB) 1.19 MB (+492.3 KB) 1.60 MB (+684.9 KB)
host/assets/truapi_verifiable_bg.wasm 4.89 MB (+703 B) 4.61 MB (-9 B) 4.64 MB (+128 B)
sandbox/polkavm-runtime/polkavm-browser-runtime.wasm 1.07 MB 186.3 KB 250.6 KB
Total 16.96 MB (+3.76 MB) 7.69 MB (+936.9 KB) (compression rate: 55%) 8.65 MB (+1.22 MB)
All files
File Raw Brotli Gzip
host/.well-known/apple-app-site-association 505 B 505 B 505 B
host/.well-known/assetlinks.json 2.1 KB 343 B 427 B
host/assets/AccountContent.js 2.0 KB (+551 B) 902 B (+164 B) 1.0 KB (+183 B)
host/assets/AuthModal.js 5.1 KB 2.0 KB (-2 B) 2.3 KB
host/assets/Chain.js 29.3 KB (+6.0 KB) 7.8 KB (+1.6 KB) 8.6 KB (+1.8 KB)
host/assets/ChainsContent.js 6.0 KB 2.2 KB (+4 B) 2.5 KB (+2 B)
host/assets/ChatDock.js 2.3 KB (+148 B) 1.1 KB (+71 B) 1.2 KB (+78 B)
host/assets/ChatPanel.js 21.8 KB (+2.5 KB) 6.8 KB (+732 B) 7.6 KB (+806 B)
host/assets/Landing.js 7.1 KB 2.8 KB (+2 B) 3.3 KB (+1 B)
host/assets/LandingPage.js 2.0 KB (+189 B) 914 B (+74 B) 1022 B (+85 B)
host/assets/LoadingScreen.js 3.8 KB 1.5 KB (+2 B) 1.8 KB (-3 B)
host/assets/PermissionsContent.js 9.2 KB (+276 B) 2.6 KB (+35 B) 3.0 KB (+35 B)
host/assets/Popover.js 4.9 KB 1.9 KB (-3 B) 2.2 KB (-2 B)
host/assets/SessionStore.js 49.9 KB 14.7 KB 18.1 KB
host/assets/SettingsContent.js 8.3 KB (+395 B) 2.9 KB (+105 B) 3.3 KB (+117 B)
host/assets/ThemeToggle.js 9.5 KB (+894 B) 2.9 KB (+296 B) 3.3 KB (+337 B)
host/assets/Topbar.astro_astro_type_script_index_0_lang.js 591 B 591 B 591 B
host/assets/TopbarActions.js 9.9 KB (+36 B) 3.3 KB (+15 B) 3.7 KB (+23 B)
host/assets/TopbarReveal.js 831 B 831 B 831 B
host/assets/UrlPill.astro_astro_type_script_index_0_lang.js 526 B 526 B 526 B
host/assets/UrlPillShield.js 2.2 KB 951 B (+2 B) 1.0 KB (+2 B)
host/assets/VerificationContent.js 1.6 KB 533 B (-1 B) 633 B
host/assets/_md.js 1.6 KB (-2 B) 799 B (+1 B) 867 B (-1 B)
host/assets/_u64.js 768 B 768 B 768 B
host/assets/account.js 990 B (+194 B) 990 B (+194 B) 990 B (+194 B)
host/assets/active.CPsmhfHv.js 238 B 238 B 238 B
host/assets/app-roots.js 314 B 314 B 314 B
host/assets/auth-controller.js 5.1 KB (+317 B) 1.9 KB (+93 B) 2.2 KB (+105 B)
host/assets/auth.js 372 B 372 B 372 B
host/assets/base.js 4.6 KB 1.9 KB 2.0 KB
host/assets/blake2.js (×2) 12.7 KB (-480 B) 4.9 KB (-182 B) 5.6 KB (-180 B)
host/assets/block-watch.js 1.1 KB (-5 B) 565 B (-6 B) 642 B (-5 B)
host/assets/bridge.js 793.8 KB (+696.8 KB) 173.1 KB (+151.6 KB) 222.8 KB (+196.7 KB)
host/assets/browser.js 22.6 KB 7.4 KB (-1 B) 8.4 KB (-1 B)
host/assets/chain-sync.js 3.1 KB (+238 B) 1.2 KB (+87 B) 1.4 KB (+112 B)
host/assets/chat-panel.js 2.7 KB (-5 B) 1.0 KB (+4 B) 1.1 KB (+1 B)
host/assets/cid-cache.js 3.1 KB (+1015 B) 1.0 KB (+255 B) 1.2 KB (+296 B)
host/assets/client.js (×2) 13.1 KB (+1.7 KB) 4.5 KB (+456 B) 5.1 KB (+523 B)
host/assets/create-popover.js 4.6 KB 1.6 KB (+1 B) 1.8 KB (-1 B)
host/assets/create-store.js 106.1 KB 31.7 KB (-17 B) 35.5 KB (+5 B)
host/assets/db.js 1.9 KB (-18 B) 677 B 798 B
host/assets/device.js 235 B 235 B 235 B
host/assets/dist.js 65 B 65 B 65 B
host/assets/dotli-debug-bus.js (×2) 1.1 KB (+367 B) 1.1 KB (+367 B) 1.1 KB (+367 B)
host/assets/drag.js 430 B 430 B 430 B
host/assets/errors.js (×2) 3.1 KB 1.6 KB 1.7 KB
host/assets/fetch.js 64.6 KB (+203 B) 16.7 KB (+82 B) 19.2 KB (+50 B)
host/assets/focus.js 1.0 KB 435 B 534 B
host/assets/hex.js 160 B 160 B 160 B
host/assets/html.js 574 B 574 B 574 B
host/assets/idle.js 323 B 323 B 323 B
host/assets/index.css 66.4 KB (+2.1 KB) 10.7 KB (+339 B) 12.2 KB (+455 B)
host/assets/index.astro_astro_type_script_index_0_lang.js 42.2 KB (+2.4 KB) 13.5 KB (+853 B) 15.0 KB (+930 B)
host/assets/ipfs.js 789 B 789 B 789 B
host/assets/lazy.js (×2) 11.3 KB (+881 B) 4.5 KB (+681 B) 4.9 KB (+718 B)
host/assets/loading.js 242 B 242 B 242 B
host/assets/log.js 3.8 KB (+227 B) 1.5 KB (+101 B) 1.7 KB (+104 B)
host/assets/manifest-types.js 7.1 KB (+3.4 KB) 2.2 KB (+1.0 KB) 2.4 KB (+1.1 KB)
host/assets/manifest.js 22.2 KB (+1.1 KB) 7.2 KB (+314 B) 8.0 KB (+378 B)
host/assets/mode.js 2.0 KB (+193 B) 713 B (+71 B) 787 B (+85 B)
host/assets/mount.js (×2) 145.4 KB (+37.2 KB) 41.1 KB (+9.6 KB) 47.4 KB (+11.2 KB)
host/assets/mount.css 22.8 KB 4.7 KB 5.3 KB
host/assets/network.js (×2) 6.4 KB (+237 B) 2.5 KB (+99 B) 2.9 KB (+95 B)
host/assets/perf.js 148 B 148 B 148 B
host/assets/permission-changes.js 342 B 342 B 342 B
host/assets/permissions.js (×2) 2.4 KB (+157 B) 1.0 KB (+49 B) 1.1 KB (+60 B)
host/assets/preload-helper.js 1.3 KB 606 B 731 B
host/assets/product-frame-layout.js 1.2 KB 548 B (+2 B) 605 B
host/assets/product.js 412 B 412 B 412 B
host/assets/proofs.js 26.0 KB 7.5 KB (-8 B) 8.3 KB (+1 B)
host/assets/provider.js 22.2 KB 5.9 KB (+5 B) 6.7 KB (+3 B)
host/assets/recent-labels.js 4.9 KB 2.1 KB (-1 B) 2.4 KB (+3 B)
host/assets/resolve.js 3.9 KB (-466 B) 1.5 KB (-208 B) 1.7 KB (-220 B)
host/assets/rolldown-runtime.js 898 B (+182 B) 898 B (+182 B) 898 B (+182 B)
host/assets/root.js 441 B 441 B 441 B
host/assets/rpc-chain.js 28.9 KB 9.3 KB (+3 B) 10.2 KB (+4 B)
host/assets/rpc-resolve.js 2.7 KB (+240 B) 1.1 KB (+81 B) 1.2 KB (+87 B)
host/assets/runtime-config.js 13.0 KB (+8.3 KB) 4.5 KB (+3.0 KB) 5.0 KB (+3.3 KB)
host/assets/scale-ts.js 4.8 KB 1.9 KB 2.1 KB
host/assets/scale.js 1.6 KB (-559 B) 740 B (-259 B) 849 B (-277 B)
host/assets/scheduled-notifications.js 8.7 KB 3.0 KB 3.3 KB
host/assets/settings.js 399 B (+59 B) 399 B (+59 B) 399 B (+59 B)
host/assets/sha2.js 1.9 KB 968 B 1.1 KB
host/assets/shared-mode.js (×2) 2.0 KB (+36 B) 975 B (+99 B) 1.0 KB (+100 B)
host/assets/solid.js 48.0 KB 16.5 KB 18.0 KB
host/assets/spans.js 741 B (-24 B) 741 B (-24 B) 741 B (-24 B)
host/assets/src.js 79.9 KB (+32 B) 25.1 KB (+59 B) 27.8 KB (+17 B)
host/assets/theme-controller.js 801 B 801 B 801 B
host/assets/toasts.js 4.3 KB (+7 B) 1.6 KB (+4 B) 1.8 KB (+3 B)
host/assets/topbar.DSRZOFUa.js 2.3 KB 950 B 1.1 KB
host/assets/topbar.Bq-n4Klp.js 196 B 196 B 196 B
host/assets/topbar.js 620 B 620 B 620 B
host/assets/truapi_provider_bg.wasm 4.01 MB 955.7 KB 1.27 MB
host/assets/truapi_server_bg.wasm 4.32 MB (+1.63 MB) 1.19 MB (+492.3 KB) 1.60 MB (+684.9 KB)
host/assets/truapi_verifiable.js 14.3 KB 2.8 KB 3.2 KB
host/assets/truapi_verifiable_bg.wasm 4.89 MB (+703 B) 4.61 MB (-9 B) 4.64 MB (+128 B)
host/assets/twoX.js 8.5 KB 3.0 KB (+1 B) 3.3 KB
host/assets/ui.js 4.1 KB 1.4 KB (-2 B) 1.6 KB
host/assets/url-pill.js 371 B 371 B 371 B
host/assets/use-store.js 368 B 368 B 368 B
host/assets/utils.js 3.5 KB 1.4 KB 1.5 KB
host/assets/verification-glyphs.js 259 B 259 B 259 B
host/assets/verify.js (×2) 12.7 KB (-1.9 KB) 4.3 KB (-884 B) 4.7 KB (-1.2 KB)
host/assets/web.js 24.9 KB (-24.6 KB) 8.1 KB (-6.5 KB) 8.9 KB (-7.3 KB)
host/assets/worker-runtime.js 52.9 KB (+17.0 KB) 11.9 KB (+3.3 KB) 13.6 KB (+3.8 KB)
host/dotli.png 11.5 KB 11.5 KB 11.5 KB
host/favicon.svg 1.8 KB 1.8 KB 1.8 KB
host/host-sw.js 7.1 KB (+106 B) 2.1 KB (+36 B) 2.4 KB (+52 B)
host/host_version.json 109 B 109 B 109 B
host/icon-192.png 12.5 KB 12.5 KB 12.5 KB
host/icon-512.png 42.8 KB 42.8 KB 42.8 KB
host/index.html 33.3 KB (+529 B) 9.0 KB (+109 B) 10.8 KB (+141 B)
host/manifest.webmanifest 441 B 441 B 441 B
host/workbox.js 14.8 KB 4.6 KB 5.1 KB
sandbox/app-sw.js 6.0 KB (+1.2 KB) 2.2 KB (+388 B) 2.5 KB (+448 B)
sandbox/assets/bitswap-bridge.js 1.0 KB 555 B (-8 B) 621 B
sandbox/assets/fetch.js 3.7 KB 1.3 KB (+4 B) 1.5 KB (-3 B)
sandbox/assets/index.css 66.4 KB (+2.1 KB) 10.7 KB (+339 B) 12.2 KB (+455 B)
sandbox/assets/index.js 313.0 KB (+81.9 KB) 87.8 KB (+22.1 KB) 102.0 KB (+25.6 KB)
sandbox/assets/mount.js 53.7 KB (+449 B) 18.2 KB (+126 B) 20.0 KB (+133 B)
sandbox/favicon.svg 1.8 KB 1.8 KB 1.8 KB
sandbox/host_version.json 108 B 108 B 108 B
sandbox/index.html 2.6 KB (+72 B) 774 B (+24 B) 1.0 KB (+35 B)
sandbox/polkavm-runtime/LICENSE-MPL-2.0 16.3 KB 16.3 KB 16.3 KB
sandbox/polkavm-runtime/PolkaVM-LICENSE-APACHE 10.6 KB 10.6 KB 10.6 KB
sandbox/polkavm-runtime/PolkaVM-LICENSE-MIT 1023 B 1023 B 1023 B
sandbox/polkavm-runtime/SHA256SUMS 631 B 631 B 631 B
sandbox/polkavm-runtime/SOURCE.json 515 B 515 B 515 B
sandbox/polkavm-runtime/polkavm-browser-runtime.wasm 1.07 MB 186.3 KB 250.6 KB
sandbox/polkavm-runtime/polkavm-gpu-worker.js 67.5 KB 11.5 KB 13.4 KB
sandbox/polkavm-runtime/polkavm-worker.js 116.9 KB 18.6 KB 22.6 KB
Total 16.96 MB (+3.76 MB) 7.69 MB (+936.9 KB) (compression rate: 55%) 8.65 MB (+1.22 MB)

Commit: 51097d6

@github-actions

github-actions Bot commented Sep 17, 2026 •

Copy link
Copy Markdown
Contributor

⚡ Performance Report

Cold Start

Overall: 🟠 uncertain  |  p50: 11.30s → 12.12s (+7.2%)  |  p95: 14.48s → 13.06s (-9.9%)  |  Mann-Whitney: z=1.77 not significant

Phase breakdown
Phase Base p50 PR p50 p50 Δ Base p95 PR p95 p95 Δ Verdict
End-to-end 11.30s 12.12s +816ms (+7.2%) 14.48s 13.06s -1427ms (-9.9%) 🟠 uncertain
Host total 6.26s 9.01s +2.75s (+43.9%) 6.58s 10.03s +3.45s (+52.5%) 🔴 regressed
Name resolution 6.02s 8.58s +2.56s (+42.4%) 6.33s 9.54s +3.22s (+50.8%) 🔴 regressed
App total 3.09s 2.92s -168ms (-5.4%) 4.05s 3.18s -863ms (-21.3%) 🟠 uncertain

Warm Start

Overall: 🔴 regressed  |  p50: 477ms → 8.84s (+1752.2%)  |  p95: 518ms → 10.17s (+1863.9%)  |  Mann-Whitney: z=4.40 ✅ significant

Phase breakdown
Phase Base p50 PR p50 p50 Δ Base p95 PR p95 p95 Δ Verdict
End-to-end 477ms 8.84s +8.36s (+1752.2%) 518ms 10.17s +9.65s (+1863.9%) 🔴 regressed
Host total 149ms 8.50s +8.35s (+5603.4%) 176ms 9.84s +9.67s (+5492.0%) 🔴 regressed
Name resolution — 8.31s new — — — —
App total 299ms 309ms +10ms (+3.3%) 334ms 325ms -9ms (-2.7%) 🔴 regressed

Lukewarm Start (different site, same session)

Overall: 🔴 regressed  |  p50: 11.87s → 15.30s (+28.9%)  |  p95: 19.36s → 15.86s (-18.1%)  |  Mann-Whitney: z=2.90 ✅ significant

Phase breakdown
Phase Base p50 PR p50 p50 Δ Base p95 PR p95 p95 Δ Verdict
End-to-end 11.87s 15.30s +3.43s (+28.9%) 19.36s 15.86s -3501ms (-18.1%) 🔴 regressed
Host total 7.11s 10.37s +3.26s (+45.9%) 7.42s 11.21s +3.79s (+51.0%) 🔴 regressed
Name resolution 5.79s 8.17s +2.38s (+41.2%) 6.09s 9.03s +2.94s (+48.3%) 🔴 regressed
App total 4.25s 4.65s +401ms (+9.4%) 5.88s 5.25s -623ms (-10.6%) 🟠 uncertain

Commit: 51097d6  |  Outliers (>2x best) discarded before stats

…6-final

# Conflicts:
#	vendor/truapi-host.lock.json
#	vendor/truapi-host/dist/wasm/web/truapi_server.d.ts
#	vendor/truapi-host/dist/wasm/web/truapi_server.js
#	vendor/truapi-host/dist/wasm/web/truapi_server_bg.wasm
#	vendor/truapi-host/dist/wasm/web/truapi_server_bg.wasm.br
#	vendor/truapi-host/dist/wasm/web/truapi_server_bg.wasm.d.ts
#	vendor/truapi-host/dist/wasm/web/truapi_server_bg.wasm.gz

This branch had an error being deployed

1 failed (outdated) deployment
paseo.fyi — b68d9757 Deployed Sep 26, 2026 by replghost via Deploy paseo.fyi #1387
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant