docs: lead README benchmarks with LexBench WebFetch

[lanyue-llk]

Problem

The README leads with the historical 192-URL crawl sample. The reviewed LexBench WebFetch comparison is missing from the project overview.

Change

Replace that sample with the public five-browser, 430-task table. Show one Moli result, task counts, pass rates and the full report link. The native Markdown disclosure opens by default. English, Chinese, German, Spanish, French, Japanese and Korean use the same measurements.

This PR targets main directly and changes only seven README files. It does not include #1380's WebMainBench table, benchmark reorganization or runtime changes. Other existing benchmark sections remain unchanged.

Reproduce and validate

Open this branch's README at the benchmark section. The LexBench WebFetch disclosure should already be open, with five rows and exactly one Moli row: 385/430 (89.53%). The report link opens the complete task-level results. Click the disclosure or focus its summary and press Enter to close or reopen it.

All seven editions pass GitHub Markdown rendering checks for the same values, five rows, one Moli result and the default open state. No WebMainBench table or internal build history remains in this change. The result counts are copied from the published report; no evaluation calls or measured evidence were changed. git diff --check passes.

Report: LexBench WebFetch.

[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 3c0cdbc5c0b5 → HEAD 8ca9e66a37c0; benchmark order: head-first; workflow: success.

Public 48-site run · informational

Metric	Base	HEAD	Δ
Run status	success	success	—
Extracted rows / contract	179 / 240	184 / 240	+5 (+2.79%)
Contract fill	74.58%	76.67%	+2.09 pp (+2.80%)
Sites with rows	37 / 48	38 / 48	+1 (+2.70%)
Sites meeting row contract	35 / 48	36 / 48	+1 (+2.86%)
Unexpected empty sites	11	10	-1 (-9.09%)
Partial-row sites	2	2	0 (0.00%)
Site outcomes	extracted: 36; HTTP error + rows: 1; empty: 7; HTTP error + empty: 1; navigation error + empty: 1; snapshot error + empty: 2	extracted: 36; HTTP error + rows: 2; empty: 7; HTTP error + empty: 1; snapshot error + empty: 2	—
Observed duration	57.11 s	50.24 s	-6.88 s (-12.04%)
Estimated CPU	50.65 s	48.14 s	-2.51 s (-4.96%)
Peak PSS	286.43 MiB	283.83 MiB	-2.59 MiB (-0.91%)
Peak RSS	290.05 MiB	287.30 MiB	-2.75 MiB (-0.95%)
Peak threads	26	27	+1 (+3.85%)
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	25 / 40	30 / 40	+5

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" [1.537, 3.539, 5.041, 5.541, 6.542, 8.045, 10.048, 11.55, 13.052, 15.054, 16.055, 16.556, 18.058, 19.56, 21.563, 23.065, 24.567, 25.569, 26.57, 28.073, 29.576, 31.078, 33.081, 34.583, 36.083, 38.085, 39.587, 41.089, 42.589, 44.593, 46.093, 47.593, 49.593, 51.096, 52.596, 54.099, 56.102, 57.103]
    y-axis "CPU percent" 0 --> 500
    line [113.71, 135.94, 159.91, 199.98, 151.75, 115.81, 135.83, 141.79, 149.83, 2, 0, 123.9, 153.83, 2, 133.96, 115.85, 139.87, 107.75, 141.96, 107.75, 145.81, 129.81, 139.8, 109.89, 28.01, 3.99, 121.86, 115.87, 8, 5.99, 0, 6, 7.99, 0, 126.02, 167.73, 109.74, 147.95]

Loading

xychart-beta
    title "Base memory: RSS"
    x-axis "Elapsed seconds" [0.036, 1.537, 3.539, 5.041, 5.541, 6.542, 8.045, 10.048, 11.55, 13.052, 15.054, 16.055, 16.556, 18.058, 19.56, 21.563, 23.065, 24.567, 25.569, 26.57, 28.073, 29.576, 31.078, 33.081, 34.583, 36.083, 38.085, 39.587, 41.089, 42.589, 44.593, 46.093, 47.593, 49.593, 51.096, 52.596, 54.099, 56.102, 57.103, 57.115]
    y-axis "Memory MiB" 0 --> 500
    line [12.66, 103.22, 126.41, 158.79, 174.37, 190.03, 198.51, 190, 176.86, 177.07, 155.6, 155.35, 162.09, 177.79, 148.23, 183.14, 198.02, 165.76, 290.05, 204.61, 226.2, 159.3, 185.58, 172.97, 166.62, 152.78, 154.7, 242.25, 226.24, 181.96, 179.95, 179.96, 179.95, 179.95, 179.95, 170.43, 195.16, 191.57, 0, 0]

Loading

HEAD

xychart-beta
    title "HEAD CPU"
    x-axis "Elapsed seconds" [1.535, 2.538, 3.039, 4.04, 5.543, 7.044, 8.547, 10.048, 11.049, 12.55, 14.053, 15.556, 17.057, 17.558, 18.058, 19.56, 21.063, 22.564, 24.067, 25.568, 26.57, 28.071, 29.573, 31.075, 32.578, 33.579, 35.08, 36.584, 38.084, 39.587, 40.587, 42.089, 43.592, 45.093, 46.596, 47.097, 47.598, 49.099]
    y-axis "CPU percent" 0 --> 500
    line [113.96, 111.7, 163.73, 113.89, 183.57, 153.94, 121.99, 173.8, 139.76, 109.85, 103.8, 71.87, 5.99, 0, 4, 101.89, 17.96, 2, 5.99, 61.89, 69.92, 115.88, 103.81, 99.97, 97.97, 129.81, 71.85, 79.85, 75.93, 7.99, 144.03, 91.85, 7.99, 79.89, 111.75, 209.58, 27.95, 111.96]

Loading

xychart-beta
    title "HEAD memory: RSS then PSS"
    x-axis "Elapsed seconds" [0.035, 1.535, 2.538, 3.039, 4.04, 5.543, 7.044, 8.547, 10.048, 11.049, 12.55, 14.053, 15.556, 17.057, 17.558, 18.058, 19.56, 21.063, 22.564, 24.067, 25.568, 26.57, 28.071, 29.573, 31.075, 32.578, 33.579, 35.08, 36.584, 38.084, 39.587, 40.587, 42.089, 43.592, 45.093, 46.596, 47.097, 47.598, 49.099, 50.236]
    y-axis "Memory MiB" 0 --> 500
    line [12.65, 114.92, 123.57, 137.32, 142.02, 152.49, 190.02, 159.36, 187.13, 170.34, 200.52, 173.02, 166.87, 153.19, 153.47, 152.91, 166.61, 150.9, 150.64, 149.59, 165.59, 174.74, 174.11, 287.3, 211.3, 159.82, 163.07, 193.56, 154.12, 168.17, 151.09, 195.28, 173.58, 151.01, 165.36, 184.88, 168.92, 165.88, 167.46, 0]
    line [9.24, 111.21, 119.87, 133.44, 138.34, 149.13, 186.47, 155.79, 183.58, 166.77, 196.95, 169.46, 163.26, 149.58, 149.87, 149.3, 163.14, 147.3, 147.04, 145.99, 162.22, 171.16, 170.35, 283.83, 207.7, 156.28, 159.36, 189.89, 150.69, 164.68, 147.5, 186.54, 170, 147.42, 161.74, 181.34, 165.82, 162.21, 163.78, 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	19.91 s	20.01 s	+0.10 s (+0.52%)
Estimated CPU	0.48 s	0.44 s	-0.04 s (-7.90%)
Peak PSS	66.10 MiB	66.18 MiB	+0.08 MiB (+0.12%)
Peak RSS	70.16 MiB	70.18 MiB	+0.01 MiB (+0.02%)
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 0.00%
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 0.00%

Package and image size

Metric	Base	HEAD	Delta	Delta %
Raw binary	153.04 MiB	153.04 MiB	0 B	0.00%
Stripped binary	106.36 MiB	106.36 MiB	0 B	0.00%
gzip binary	52.71 MiB	52.71 MiB	−1 B	−0.000002%
Rootfs	159.94 MiB	159.94 MiB	0 B	0.00%
gzip rootfs	61.75 MiB	61.75 MiB	−2 B	−0.000003%

Startup latency and PSS

Case	Base p50	HEAD p50	Delta	Base PSS	HEAD PSS	Delta
Serve ready	3.10 ms	27.27 ms	+24.17 ms	9.66 MiB	4.32 MiB	−5.34 MiB
CDP first page	65.38 ms	50.59 ms	−14.78 ms	9.12 MiB	4.34 MiB	−4.78 MiB
CDP warm page	17.14 ms	12.35 ms	−4.79 ms	50.27 MiB	50.65 MiB	+381.00 KiB
CLI about:blank	27.52 ms	15.66 ms	−11.86 ms	6.53 MiB	4.36 MiB	−2.17 MiB
CLI local JS	18.23 ms	19.47 ms	+1.24 ms	4.33 MiB	3.71 MiB	−636.50 KiB

Concurrency matrix

Concurrency	Base p50	HEAD p50	Delta	Base PSS	HEAD PSS	Delta	HEAD failures	Unstable B/H
1	17.51 ms	18.06 ms	+0.55 ms	747.00 KiB	825.00 KiB	+78.00 KiB	0	0/0
5	20.98 ms	21.68 ms	+0.69 ms	28.43 MiB	18.67 MiB	−9.77 MiB	0	0/0
10	19.77 ms	20.78 ms	+1.00 ms	23.61 MiB	27.92 MiB	+4.31 MiB	0	0/0
25	19.68 ms	21.47 ms	+1.79 ms	13.47 MiB	28.18 MiB	+14.71 MiB	0	0/0
100	20.64 ms	21.10 ms	+0.46 ms	13.64 MiB	27.80 MiB	+14.16 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	6m 42.88s

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	4.18 ms	26.60 ms
Chromium	8	8	47/47	275.19 ms	93.57 ms

Operation p50	Chromium	Moli	Delta	Delta %
navigate	44.96 ms	5.76 ms	−39.21 ms	−87.20%
observe	0.89 ms	0.95 ms	+0.06 ms	+6.36%
fill	3.98 ms	3.61 ms	−0.38 ms	−9.46%
click	6.14 ms	4.36 ms	−1.78 ms	−28.97%

Idle-resource episode	Peak RSS	Peak PSS	Average CPU	Peak CPU	Peak processes	Sampler
Moli	66.65 MiB	63.96 MiB	0.74%	10.00%	1	✅
Chromium	1287.96 MiB	335.96 MiB	2.00%	93.72%	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	18.68 ms	20.55 ms	647.00 KiB
Long-lived CDP session	13	0	19.58 ms	20.43 ms	—

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

Workers: 4 · cumulative group time: 196.63 s · failed groups: 0

Slowest group	Scenarios	Duration	Status
scrollbar-visibility	22	28.59 s	✅
target-lifecycle	24	17.04 s	✅
dom-input	27	13.11 s	✅
multi-page	45	11.46 s	✅
chromium-cdp	56	11.45 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.

[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 3c0cdbc5c0 → HEAD 8ca9e66a37; benchmark order: head-first; workflow: success.

Session resilience

Metric	Base	HEAD
Attempted / planned	200 / 200	200 / 200
Direct observable pass	199 / 200	200 / 200
Failures / recovered	1 / 1	0 / 0
Recovery failures	0	0
Lifecycle order violations	0	0
Network order violations	0	0
Wall time	439.8 s	461.0 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	263.3 MiB	242.2 MiB	-21.0 MiB
Periodic peak PSS	260.5 MiB	240.2 MiB	-20.3 MiB
RSS · first 10 avg	177.6 MiB	173.9 MiB	-3.8 MiB
RSS · last 10 avg	196.2 MiB	185.9 MiB	-10.3 MiB
RSS · last − first window	+18.6 MiB	+12.0 MiB	-6.6 MiB
RSS · warm slope / 100 nav	+3.2 MiB	+3.2 MiB	-0.0 MiB
PSS · first 10 avg	174.4 MiB	170.9 MiB	-3.5 MiB
PSS · last 10 avg	193.5 MiB	182.8 MiB	-10.7 MiB
PSS · last − first window	+19.1 MiB	+11.9 MiB	-7.2 MiB
PSS · warm slope / 100 nav	+3.5 MiB	+3.8 MiB	+0.3 MiB
Peak file descriptors	58	62	4
Peak threads	29	31	2

HEAD memory by 50-navigation quarter

HEAD range	Avg RSS	Final RSS	Peak RSS	Avg PSS	Final PSS	Peak FD
1–50	182.6 MiB	165.2 MiB	233.6 MiB	179.3 MiB	161.2 MiB	54
51–100	182.9 MiB	138.5 MiB	237.3 MiB	179.2 MiB	134.9 MiB	55
101–150	189.0 MiB	171.7 MiB	237.8 MiB	186.0 MiB	168.8 MiB	55
151–200	188.1 MiB	175.5 MiB	241.9 MiB	185.1 MiB	172.3 MiB	54

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.