Lava lamp v1 · Claude Code · run dc76fa3b · 30.09.2026 19:51 Tallinn

Claude Opus 5.5

Low effort

Works

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

Compare →Transcript Source

Desktop capture of Claude Code (Opus 5.5)'s output on Lava lamp v1, low effort, 30.09.2026 19:51 (run dc76fa3b) — **Desktop** · 353.0 KB · e3834736a85b…

Phone capture of Claude Code (Opus 5.5)'s output on Lava lamp v1, low effort, 30.09.2026 19:51 (run dc76fa3b) — **Phone** · 166.1 KB · 8a65bf32ad02…

Motion capture of Claude Code (Opus 5.5)'s output on Lava lamp v1, low effort, 30.09.2026 19:51 (run dc76fa3b) — **Motion** · 354.7 KB · e4dad4a90faa…

Build time
1m 42s
5th fastest of 12 models at low effort. Wall time, start to graded: 2m 43s.
Output tokens
10,631
9th fewest of 12 models at low effort.
Output speed
104 tok/s
5th highest of 12 models at low effort.
API cost
$0.33
The CLI’s list-price estimate, not a charge.

Along the row

Claude Opus 5.5 at every effort.

Down the column

Other models at low effort.

Fastest median first · this run: 1m 42s

Hidden checks

8 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 6e5ad1e5829a92c0617559acd07b37fb324dc005add65a40eeeaf3a9fe75a912, 698.8 KB.

Run ID: dc76fa3b-2090-4e2a-8937-3f225de0a630
Benchmark: lava-lamp v1
Experiment: Effort pair
Started: 30.09.2026 19:51 Tallinn
Graded: revision 1 · 30.09.2026 19:54

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	pass
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-opus-5-5	claude-opus-5-5	harness result, 4 of 4 API requests
Effort	low	low	4 of 4 API requests
Harness	Claude CLI	2.1.285 (Claude Code)	Version as the CLI reported it; binary SHA-256 51f09bd1e021…
Deadline	10 min	2m 43s used	Completed

Tool decisions

3 calls · 3 accepted · 0 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.

Bash2 calls · 2 accepted · 2 failed
Write1 call · 1 accepted

Telemetry

Time and usage.

Generation time: 1m 42s
First content: 1.8 s
Output speed: 104.4 tokens/s, whole run including tools
API-equivalent cost: $0.33 (list-price estimate, not a charge)
Actual charge: not reported
Billing: subscription

Input tokens

65,532 tokens

Cache reads51,88579%
Cache writes13,63921%
Uncached input8<0.1%

Output tokens

10,631 tokens

Reasoning3,51233%
Other output7,11967%

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: 17.2 KB, 2 redactions. Download transcript.txt

Source files

1 file · 11.3 KB

Files in the output bundle
File	Size	SHA-256	Redactions
index.html	11.3 KB	42955cae1d677505b5af4ac6ae259f099e3547fb58d7340b6b6fc94a07c63881	0

Bundle SHA-256 42955cae1d677505b5af4ac6ae259f099e3547fb58d7340b6b6fc94a07c63881

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)	30.09.2026 19:54	works

Redaction: automated scan run; checked by Jaanus on 22.10.2026 15:37. Redacted: 2 local-path.

Download run.json