Lava lamp v1 · Claude Code · run 4c8fd1ea · 28.10.2026 07:54 Tallinn

Claude Sonnet 5.5

High effort

Works

7 of 8 checks passed. All 5 blocker checks passed.

Compare →Transcript Source

Desktop capture of Claude Code (Sonnet 5.5)'s output on Lava lamp v1, high effort, 28.10.2026 07:54 (run 4c8fd1ea) — **Desktop** · 704.5 KB · 7993b7035f3f…

Phone capture of Claude Code (Sonnet 5.5)'s output on Lava lamp v1, high effort, 28.10.2026 07:54 (run 4c8fd1ea) — **Phone** · 214.4 KB · 6d7f3b29764b…

Motion capture of Claude Code (Sonnet 5.5)'s output on Lava lamp v1, high effort, 28.10.2026 07:54 (run 4c8fd1ea) — **Motion** · 700.6 KB · 67df8a6f1996…

Build time
6m 45s
9th fastest of 12 models at high effort. Wall time, start to graded: 7m 43s.
Output tokens
52,729
9th fewest of 12 models at high effort.
Output speed
130 tok/s
2nd highest of 12 models at high effort.
API cost
$0.71
The CLI’s list-price estimate, not a charge.

Along the row

Claude Sonnet 5.5 at every effort.

Down the column

Other models at high effort.

Fastest median first · this run: 6m 45s

Hidden checks

7 of 8 checks passed.

Self-contained index.html is presentblocker
Visible page contentblocker
Visible autonomous animationblocker
Accessible on/off controlextra
390 px mobile layout has no horizontal overflowextra
Local animation frame performanceextra
No browser errorsblocker
No attempted external or linked-resource requestsblocker

It works when every blocker passes. How it looks is reviewed blind: pending.

Run detailsagent config, what ran, tokens, source and revisions

Video (video): not published, videos are withheld. SHA-256 ea9e5457064e991b0bf44ba5284d21ecadbbf1a12180927b1d6883a38b90ecb4, 1508.3 KB.

Run ID: 4c8fd1ea-b0b1-47c1-bfe4-fa16a2eab716
Benchmark: lava-lamp v1
Experiment: Single run
Started: 28.10.2026 07:54 Tallinn
Graded: revision 1 · 28.10.2026 08:01

Works and checks, in full

All 5 blocker checks passed.

The rule: a run works when every blocker check passes. If any blocker fails it does not work. If none fails but one has no result yet, it stays pending, never counted as a pass or a fail. Extra checks are reported but do not decide it. Works is not a judgement of looks: that is the visual review.

This is a legacy contract v1 run: the task has no public requirements, and these are the original lava-lamp checks, not a versioned hidden verifier. It is never compared with contract v2 runs.

Contract v1 checks for this run, with blocker and result
Requirement	Check	Blocker	Result
none (v1)	Self-contained index.html is present	Blocker	pass
none (v1)	Visible page content	Blocker	pass
none (v1)	Visible autonomous animation	Blocker	pass
none (v1)	Accessible on/off control	Extra	pass
none (v1)	390 px mobile layout has no horizontal overflow	Extra	fail
none (v1)	Local animation frame performance	Extra	pass
none (v1)	No browser errors	Blocker	pass
none (v1)	No attempted external or linked-resource requests	Blocker	pass

Verification completed; eligible for benchmark results.

Visual review

Pending.

Pending: the captures have not been reviewed blind yet. Checks decide whether it works; how it looks is decided only in that review.

What ran

Requested and detected.

Detected values come from the run’s own output, never from our settings.

Requested agent config beside what the run reported
Field	Requested	Detected	Evidence
Model	claude-sonnet-5-5	claude-sonnet-5-5	harness result, 7 of 7 API requests
Effort	high	high	7 of 7 API requests
Harness	not recorded	2.1.312 (Claude Code)	Version as the CLI reported it; binary SHA-256 6b5ec691195e…
Deadline	60 min	7m 43s used	Completed

Tool decisions

12 calls · 11 accepted · 1 rejected

Each tool call is allowed or blocked by the harness’s permission settings before it runs. Allowed is not the same as succeeded: an allowed call can still fail.

Bash6 calls · 5 accepted · 1 rejected · 6 failed
Edit5 calls · 5 accepted
Write1 call · 1 accepted

Telemetry

Time and usage.

Generation time: 6m 45s
First content: 1.5 s
Output speed: 130.3 tokens/s, whole run including tools
API-equivalent cost: $0.71 (list-price estimate, not a charge)
Actual charge: not reported
Billing: api

Input tokens

349,149 tokens

Cache reads286,97582%
Cache writes62,16018%
Uncached input14<0.1%

Output tokens

52,729 tokens

Reasoning37,21371%
Other output15,51629%

Usage as the agent reported it at the end of the run. Cache reads are part of input; reasoning is part of output.

Transcript and source

Plain-text downloads.

The redacted transcript and the frozen output bundle, with LF line endings. Hashes are of the original files.

Redacted agent transcript: 43.6 KB, 12 redactions. Download transcript.txt

Source files

1 file · 21.0 KB

Files in the output bundle
File	Size	SHA-256	Redactions
index.html	21.0 KB	d4c2b155b3444da829c2332f8935e35fcaf97418b2684c9df3f5ece467c78698	0

Bundle SHA-256 d4c2b155b3444da829c2332f8935e35fcaf97418b2684c9df3f5ece467c78698

Revision history

This page shows revision 1.

Grades are append-only: a regrade adds a revision and never rewrites an earlier one.

Grading revisions of this run
Revision	Graded	Works
1 (current)	28.10.2026 08:01	works

Redaction: automated scan run; checked by Jaanus on 28.10.2026 17:51. Redacted: 4 local-path, 8 username.

Download run.json