[lanyue-llk]

A successful Load can return a bootstrap page before a credentialed fetch callback replaces it. Markdown also omits live child-frame content, and ResizeObserver can read an old layout snapshot after a size change.

This change fixes those three paths:

• At Load, a pending credentialed fetch or XHR owned by the current top-level Document gets the existing bounded navigation grace. Ordinary polling and long timers do not extend completion. Replacement continuations retain the requested lifecycle stage and command completion when the Document changes; the ownership invariant remains enforced.
• Markdown --with-frames converts the current child Documents at their frame positions, including nested and cross-origin frames. Links and images use each Document's effective base URL, including inherited srcdoc bases and <base>. Frames with no usable content retain their source link or label.
• With real layout enabled, ResizeObserver delivery runs on the exact Document's rendering task after fresh layout publication. Callback mutations trigger another layout pass in the same bounded update, unchanged sizes are deduplicated, and same-depth feedback leaves an undelivered-notification warning after the update.

Default fetch --dump markdown continues to use the native text path without real layout. A CLI virtual-list fixture verifies that its resize-driven children appear with explicit --layout and remain absent without it. This does not improve dynamic-104 under the default text method.

Validation on ARM64 Linux, Rust 1.96.1:

• cargo fmt --all passed.
• Clippy passed for the moli, moli-renderer-v8, and moli-test-support libraries with all features and warnings denied; moli also passed with all targets.
• 13 selected nextest tests passed: credential callback navigation in HTML and Markdown, bounded non-navigation grace, unrelated polling/timers, current/nested/cross-origin frame content and URL bases, failed-frame links, resize-driven children and size deduplication, bounded feedback, the CLI layout boundary, and existing 403 navigation, timeout, same-document navigation, HTML-frame and semantic-tree regressions.

This branch starts from main at 7a36b6a63f7b8f8f9801e19959343620fba36265 and is independent of PR1476 and the content/navigation stack.

[github-actions]

Sequential Navigation Soak A/B

✅ HEAD completed the 200-navigation resilience and memory observation.

One browser process, one target, and one CDP session navigate CSDN → SegmentFault → Huaban → example.com repeatedly. This run issued 200 Page.navigate commands per binary.

Common ancestor 7a36b6a63f → HEAD a2c2ceeab3; benchmark order: base-first; workflow: success.

Session resilience

Metric	Base	HEAD
Attempted / planned	200 / 200	200 / 200
Direct observable pass	200 / 200	200 / 200
Failures / recovered	0 / 0	0 / 0
Recovery failures	0	0
Lifecycle order violations	0	0
Network order violations	0	0
Wall time	476.8 s	479.8 s

A public-page failure is reported separately from an unrecoverable session. The soak requires all navigation attempts, zero failed recovery, zero lifecycle/network ordering violations, and complete resource evidence.

Process-tree memory

Metric	Base	HEAD	Δ
Boundary samples	200	200	0
Boundary RSS samples	200	200	0
Boundary PSS samples	200	200	0
Periodic peak RSS	268.8 MiB	263.8 MiB	-5.0 MiB
Periodic peak PSS	265.6 MiB	262.2 MiB	-3.4 MiB
RSS · first 10 avg	180.9 MiB	181.1 MiB	+0.2 MiB
RSS · last 10 avg	183.1 MiB	191.4 MiB	+8.4 MiB
RSS · last − first window	+2.2 MiB	+10.4 MiB	+8.2 MiB
RSS · warm slope / 100 nav	+0.3 MiB	+3.6 MiB	+3.3 MiB
PSS · first 10 avg	178.1 MiB	178.2 MiB	+0.1 MiB
PSS · last 10 avg	180.3 MiB	188.3 MiB	+8.0 MiB
PSS · last − first window	+2.2 MiB	+10.1 MiB	+7.9 MiB
PSS · warm slope / 100 nav	+0.5 MiB	+3.6 MiB	+3.1 MiB
Peak file descriptors	59	59	0
Peak threads	31	29	-2

HEAD memory by 50-navigation quarter

HEAD range	Avg RSS	Final RSS	Peak RSS	Avg PSS	Final PSS	Peak FD
1–50	181.2 MiB	158.1 MiB	231.0 MiB	178.2 MiB	154.7 MiB	54
51–100	184.7 MiB	175.7 MiB	229.1 MiB	181.7 MiB	172.7 MiB	54
101–150	183.2 MiB	164.6 MiB	226.2 MiB	180.2 MiB	161.4 MiB	53
151–200	187.9 MiB	180.4 MiB	233.7 MiB	184.8 MiB	177.5 MiB	53

Raw reports and per-navigation boundary samples: workflow run and full sequential-navigation-soak-results artifact.

Public-site timing and memory are observational. Use the A/B deltas and trend shape as evidence, not as a deterministic performance threshold.

[github-actions]

Spider Bench A/B

✅ All benchmark browser service runs completed; results are informational.

Public HEAD: 184 / 240 rows; 38 / 48 sites produced rows; 10 unexpected empty sites.

✅ Deterministic fixture contract is clean.

Common ancestor 7a36b6a63f7b → HEAD a2c2ceeab30b; benchmark order: base-first; workflow: success.

Public 48-site run · informational

Metric	Base	HEAD	Δ
Run status	success	success	—
Extracted rows / contract	184 / 240	184 / 240	0 (0.00%)
Contract fill	76.67%	76.67%	0.00 pp (0.00%)
Sites with rows	38 / 48	38 / 48	0 (0.00%)
Sites meeting row contract	36 / 48	36 / 48	0 (0.00%)
Unexpected empty sites	10	10	0 (0.00%)
Partial-row sites	2	2	0 (0.00%)
Site outcomes	extracted: 36; HTTP error + rows: 2; empty: 7; HTTP error + empty: 1; snapshot error + empty: 2	extracted: 37; HTTP error + rows: 1; empty: 7; HTTP error + empty: 1; snapshot error + empty: 2	—
Observed duration	58.97 s	66.92 s	+7.96 s (+13.49%)
Estimated CPU	49.34 s	50.30 s	+0.96 s (+1.95%)
Peak PSS	270.76 MiB	272.88 MiB	+2.12 MiB (+0.78%)
Peak RSS	274.37 MiB	277.05 MiB	+2.68 MiB (+0.98%)
Peak threads	27	27	0 (0.00%)
Mismatched sites	0	0	0
Timeout with items	0	0	0

Category rows	Base	HEAD	Δ
news	35 / 40	35 / 40	0
stocks	20 / 40	20 / 40	0
tech	35 / 40	35 / 40	0
sports	27 / 40	27 / 40	0
games	37 / 40	37 / 40	0
life	30 / 40	30 / 40	0

Public-site content, timing, and memory are noisy. This single A/B run reports evidence only; site outcome counts explain missing rows without treating them as a deterministic regression.

CPU and memory timelines

These are bounded 40-point views of the same complete process-tree samples used by the HTML report. CPU uses 100% per occupied logical core; memory lines are RSS first and PSS second when complete PSS samples are available.

Base

xychart-beta
    title "Base CPU"
    x-axis "Elapsed seconds" [1.539, 3.041, 3.541, 5.041, 6.541, 7.042, 8.043, 10.045, 11.548, 13.049, 15.053, 16.554, 18.054, 19.557, 21.558, 23.061, 24.563, 26.066, 28.068, 29.57, 31.073, 33.075, 34.575, 36.076, 37.076, 37.576, 39.578, 41.079, 42.581, 44.583, 46.084, 47.587, 49.089, 51.092, 52.594, 54.096, 55.598, 57.6]
    y-axis "CPU percent" 0 --> 500
    line [35.92, 0, 98.03, 136.03, 161.98, 199.86, 139.92, 95.94, 175.75, 77.93, 133.82, 105.89, 8, 5.99, 2, 0, 2, 99.78, 89.97, 7.98, 7.98, 95.98, 125.95, 92.01, 112.06, 105.93, 137.92, 155.75, 121.73, 103.88, 133.84, 89.73, 143.84, 19.97, 31.96, 119.82, 8, 122.05]

Loading

xychart-beta
    title "Base memory: RSS then PSS"
    x-axis "Elapsed seconds" [0.036, 1.539, 3.041, 3.541, 5.041, 6.541, 7.042, 8.043, 10.045, 11.548, 13.049, 15.053, 16.554, 18.054, 19.557, 21.558, 23.061, 24.563, 26.066, 28.068, 29.57, 31.073, 33.075, 34.575, 36.076, 37.076, 37.576, 39.578, 41.079, 42.581, 44.583, 46.084, 47.587, 49.089, 51.092, 52.594, 54.096, 55.598, 57.6, 58.968]
    y-axis "Memory MiB" 0 --> 500
    line [12.39, 89.73, 100.05, 117.92, 153.62, 152.64, 174.47, 183.34, 168.13, 171.99, 160.74, 177.52, 144.11, 143.04, 143.96, 144.71, 144.91, 145.33, 162.89, 151.32, 136.88, 139.39, 174.06, 165.86, 170.44, 274.37, 191.67, 186.6, 160.23, 206.78, 157.21, 154.99, 178.29, 234.68, 172.89, 162.72, 184.19, 147.27, 177.9, 0]
    line [9.03, 86.02, 96.33, 114.33, 150.01, 149.03, 170.39, 179.73, 164.39, 168.22, 157.01, 173.82, 140.38, 139.31, 140.23, 140.98, 141.19, 141.61, 159.18, 147.59, 133.2, 135.67, 170.49, 162.15, 166.74, 270.76, 187.94, 182.87, 156.52, 203.16, 153.6, 151.26, 174.67, 230.96, 170.37, 159.09, 180.48, 143.54, 174.16, 0]

Loading

HEAD

xychart-beta
    title "HEAD CPU"
    x-axis "Elapsed seconds" [1.037, 1.537, 2.037, 3.54, 5.545, 7.547, 9.549, 11.051, 13.053, 15.055, 17.057, 18.559, 20.563, 22.564, 24.067, 26.069, 28.071, 30.073, 31.576, 33.578, 35.581, 37.082, 39.084, 41.086, 43.089, 44.591, 46.593, 48.594, 50.597, 52.098, 54.1, 56.103, 57.605, 59.608, 61.61, 63.614, 65.115, 65.615]
    y-axis "CPU percent" 0 --> 500
    line [0, 8, 2, 119.76, 127.67, 137.87, 91.92, 129.9, 101.92, 133.79, 6, 0, 0, 119.86, 17.98, 33.95, 11.99, 11.99, 0, 0, 27.97, 9.98, 13.99, 3.99, 93.89, 113.83, 116.01, 121.8, 173.98, 127.97, 115.9, 59.95, 119.79, 15.98, 133.99, 33.94, 132, 203.99]

Loading

xychart-beta
    title "HEAD memory: RSS then PSS"
    x-axis "Elapsed seconds" [0.036, 1.037, 1.537, 2.037, 3.54, 5.545, 7.547, 9.549, 11.051, 13.053, 15.055, 17.057, 18.559, 20.563, 22.564, 24.067, 26.069, 28.071, 30.073, 31.576, 33.578, 35.581, 37.082, 39.084, 41.086, 43.089, 44.591, 46.593, 48.594, 50.597, 52.098, 54.1, 56.103, 57.605, 59.608, 61.61, 63.614, 65.115, 65.615, 66.924]
    y-axis "Memory MiB" 0 --> 500
    line [12.29, 54.86, 63.14, 67.13, 115.43, 144.95, 171.11, 164.98, 182.79, 191.23, 166.07, 150.41, 149.98, 150.88, 178.55, 161.78, 164.11, 160.66, 162.25, 161.11, 161.44, 163.23, 161.63, 138.75, 140.69, 167.26, 180.03, 277.05, 232.34, 174.89, 174.96, 164.98, 159.59, 170.03, 167.45, 195.77, 159.25, 200.6, 216.01, 0]
    line [8.97, 51.21, 59.49, 63.5, 111.83, 141.73, 167.39, 161.3, 179.91, 188.14, 162.32, 146.68, 146.25, 147.14, 174.86, 158.05, 160.4, 156.94, 158.52, 157.38, 157.71, 159.5, 157.9, 135.03, 136.96, 163.53, 176.97, 272.88, 228.75, 171.6, 171.24, 161.33, 155.86, 166.36, 163.72, 192.05, 156.27, 196.87, 212.27, 0]

Loading

Deterministic fixture · required contract

The fixture exercises 15 routes: 8 are expected to emit rows and 7 intentionally exercise empty, loading, or timeout behavior.

Metric	Base	HEAD	Δ
Run status	success	success	—
Extracted rows / contract	40 / 40	40 / 40	0 (0.00%)
Contract fill	100.00%	100.00%	0.00 pp (0.00%)
Sites with rows	8 / 15	8 / 15	0 (0.00%)
Sites meeting row contract	15 / 15	15 / 15	0 (0.00%)
Unexpected empty sites	0	0	0
Partial-row sites	0	0	0
Site outcomes	extracted: 7; HTTP error + rows: 1; empty: 3; timeout + empty: 4	extracted: 7; HTTP error + rows: 1; empty: 3; timeout + empty: 4	—
Observed duration	19.85 s	19.86 s	+0.01 s (+0.07%)
Estimated CPU	0.40 s	0.42 s	+0.02 s (+5.22%)
Peak PSS	66.17 MiB	65.90 MiB	-0.27 MiB (-0.41%)
Peak RSS	69.66 MiB	69.36 MiB	-0.31 MiB (-0.44%)
Peak threads	18	18	0 (0.00%)
Mismatched sites	0	0	0
Timeout with items	0	0	0

Full HTML, JSON, CSV, logs, and page snapshots: workflow run and full spider-bench-results artifact.

The public-web A/B is informational. The exact HEAD fixture contract runs as a separate required CI check; timing and public-site content deltas remain non-blocking.

[github-actions]

CI Regression Report

Source CI run · source state at render: in_progress · artifacts: 6/6

Check	Status	Signal
Release regression	✅	HEAD/base failures 0/0; raw binary +0.025550%
Frontend differential	✅	1,020/1,020 cases matched; 0 issues
Agent episodes	✅	8/8 Moli episodes passed; 0 failures
Runtime/CDP contracts	✅	26 contract cases; 0 failures
CDP smoke	✅	52/52 groups passed; 621 scenarios
WebMainBench	✅	545/545 completed; 0 unexpected failures

Release regression — ✅ HEAD/base failures 0/0; raw binary +0.025550%

Package and image size

Metric	Base	HEAD	Delta	Delta %
Raw binary	150.47 MiB	150.51 MiB	+39.37 KiB	+0.025550%
Stripped binary	103.94 MiB	103.98 MiB	+36.00 KiB	+0.033823%
gzip binary	51.74 MiB	51.75 MiB	+15.02 KiB	+0.028352%
Rootfs	157.37 MiB	157.41 MiB	+40.00 KiB	+0.024822%
gzip rootfs	60.68 MiB	60.70 MiB	+20.35 KiB	+0.032748%

Startup latency and PSS

Case	Base p50	HEAD p50	Delta	Base PSS	HEAD PSS	Delta
Serve ready	27.24 ms	27.38 ms	+0.14 ms	3.81 MiB	3.99 MiB	+185.50 KiB
CDP first page	51.30 ms	50.59 ms	−0.71 ms	4.20 MiB	3.87 MiB	−334.00 KiB
CDP warm page	12.87 ms	11.94 ms	−0.92 ms	49.16 MiB	49.40 MiB	+244.50 KiB
CLI about:blank	14.86 ms	15.99 ms	+1.13 ms	2.89 MiB	2.51 MiB	−387.00 KiB
CLI local JS	19.13 ms	20.07 ms	+0.94 ms	2.63 MiB	1.40 MiB	−1.23 MiB

Concurrency matrix

Concurrency	Base p50	HEAD p50	Delta	Base PSS	HEAD PSS	Delta	HEAD failures	Unstable B/H
1	19.48 ms	19.82 ms	+0.34 ms	573.00 KiB	575.00 KiB	+2.00 KiB	0	0/0
5	22.14 ms	21.67 ms	−0.47 ms	13.36 MiB	27.23 MiB	+13.87 MiB	0	0/0
10	21.15 ms	21.72 ms	+0.57 ms	13.49 MiB	12.62 MiB	−895.00 KiB	0	0/0
25	21.72 ms	21.27 ms	−0.44 ms	22.84 MiB	22.72 MiB	−120.50 KiB	0	0/0
100	20.98 ms	22.16 ms	+1.18 ms	18.22 MiB	40.76 MiB	+22.53 MiB	0	0/0

Frontend differential — ✅ 1,020/1,020 cases matched; 0 issues

Match	DOM mismatch	Diagnostic mismatch	Moli error	Reference error	Infrastructure error	Mismatched frames	Duration
1,020	0	0	0	0	0	0	7m 35.59s

Agent episodes · Moli vs Chromium — ✅ 8/8 Moli episodes passed; 0 failures

Fast contract	Episodes	Passed	Assertions	Ready p50	Episode p50
Moli	8	8	47/47	4.48 ms	27.64 ms
Chromium	8	8	47/47	277.22 ms	96.57 ms

Operation p50	Chromium	Moli	Delta	Delta %
navigate	50.04 ms	6.61 ms	−43.43 ms	−86.79%
observe	0.86 ms	1.10 ms	+0.23 ms	+26.94%
fill	3.61 ms	3.32 ms	−0.29 ms	−8.10%
click	5.21 ms	4.85 ms	−0.36 ms	−6.99%

Idle-resource episode	Peak RSS	Peak PSS	Average CPU	Peak CPU	Peak processes	Sampler
Moli	65.69 MiB	63.10 MiB	0.78%	10.00%	1	✅
Chromium	1286.73 MiB	331.43 MiB	2.12%	102.86%	13	⚠️ PSS partial

Runtime and CDP session contracts — ✅ 26 contract cases; 0 failures

Suite	Cases	Failures	Median case p50	Median case p95	Median peak PSS p50
Synthetic fetch	13	0	18.91 ms	20.25 ms	1.03 MiB
Long-lived CDP session	13	0	18.70 ms	19.66 ms	—

CDP smoke — ✅ 52/52 groups passed; 621 scenarios

Workers: 4 · cumulative group time: 195.94 s · failed groups: 0

Slowest group	Scenarios	Duration	Status
scrollbar-visibility	22	28.11 s	✅
target-lifecycle	24	17.07 s	✅
dom-input	27	12.80 s	✅
multi-page	45	11.45 s	✅
chromium-cdp	56	10.95 s	✅

WebMainBench · 545 pages — ✅ 545/545 completed; 0 unexpected failures

Completed	Successful	Expected DNS failure	Unexpected failures	Panics	Crashes	Timeouts	Empty output
545/545	544	1	0	0	0	0	0

Frozen HTML · Linux · --wait done · 45 seconds per page · no retries. Content quality is evaluated separately.

Diagnostics: webmainbench-results — every page's Markdown, stderr, and exit status. Source run and artifacts

All artifact fields are parsed by the trusted default-branch renderer; missing or invalid inputs remain visible as unavailable.

[lanyue-llk]

Superseded by #1501 for the evidenced Web Fetch fixes: frame Markdown, per-frame stripping, navigation-stage preservation, and ResizeObserver delivery/recovery. The credential-request navigation heuristic was intentionally excluded.