Task gallery / DeepSWE (long-horizon SWE) / Version 1
Reference solution (oracle)
Unverified
No-op resistance check
Unverified
QA verdict
Not recorded
Aggregate eligibility
Eligible version

The task

ammonia sanitizes HTML through its own Builder, but has no implementation of the WICG HTML Sanitizer API: no SanitizerConfig model, no config canonicalization, validity checking or reflection, and no way to run one config in safe and unsafe modes. Implement the configuration semantics of the WICG HTML Sanitizer API (Draft Community Group Report as of 24 November 2025) as a new public module ammonia::sanitizer (new s…

Excerpt only; not a complete executable task specification. Task ID task_40df7b0ce1654f8f · version 1. Selection status: not fully recorded.

Results by model and effort

This exact task version only. Every rate is passes / valid runs with a Wilson 95% interval; n≤2 remains counts only. Macro aggregation does not apply within a single task. Baselines are summarized in the quality gates, not ranked. A passed no-op resistance check means that doing nothing did not solve the task.

Current evidence

Model / effortVerifier result [95% Wilson]Diagnostic labels / validHARNESS_ERROR · excludedOutcome classes
Claude Opus 5 · default (unverified)5/5 · 100.0% [56.6%, 100.0%]5 / 50
GOOD_FAILURE: 0
BAD_FAILURE: 0
GOOD_SUCCESS: 5
BAD_SUCCESS: 0
GPT-5.6-sol · default (unverified)5/5 · 100.0% [56.6%, 100.0%]5 / 50
GOOD_FAILURE: 0
BAD_FAILURE: 0
GOOD_SUCCESS: 5
BAD_SUCCESS: 0

Legacy evidence

Model / effortVerifier result [95% Wilson]Diagnostic labels / validHARNESS_ERROR · excludedOutcome classes
Claude Opus 4.8 · default (unverified)3/5 · 60.0% [23.1%, 88.2%]5 / 50
GOOD_FAILURE: 2
BAD_FAILURE: 0
GOOD_SUCCESS: 3
BAD_SUCCESS: 0
Gemini Flash (latest alias) · default (unverified)0/11 / 11
GOOD_FAILURE: 1
BAD_FAILURE: 0
GOOD_SUCCESS: 0
BAD_SUCCESS: 0

Difficulty, with uncertainty

Selected-best difficulty: 0.0% [0.0%, 43.4%]. This is one minus the best measured current-row pass rate among rows with at least five valid runs.

The maximum is selected from the observed rows, so the estimate is optimistic. Its Wilson interval is not adjusted for model selection. “Hard” requires qualifying GOOD_FAILURE evidence; the threshold flag additionally requires the interval and recorded quality gates to meet the documented rule. This is not independent validation. Threshold flag: no.

Every recorded run

38 recorded trials, including controls. Diagnostic labels do not replace verifier reward. Labels and quality checks are source declarations, not independently attested here. Internal evidence remains private; a missing public URL is shown explicitly. Records with invalid or absent rewards are not scored.

Run / modelDate / sourceReward / eligibilityDiagnostic class / sourceResourcesEvidence
trial_0ba3981e05384dd8
Claude Opus 4.8 · default (unverified)
Aug 29, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 84,139
Turns: 60
Cost: $7.0017
Not publicly available
trial_557a3b8dfae54cea
Claude Opus 4.8 · default (unverified)
Aug 29, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 93,309
Turns: 55
Cost: $6.2024
Not publicly available
trial_5c43c07be63741f8
Claude Opus 4.8 · default (unverified)
Aug 29, 2026
console
0 · validGOOD_FAILURE
frozen-classifier
Tokens: 80,147
Turns: 49
Cost: $6.1511
Not publicly available
trial_782c4d7bf4a64e88
Claude Opus 4.8 · default (unverified)
Aug 29, 2026
console
0 · validGOOD_FAILURE
frozen-classifier
Tokens: 88,437
Turns: 55
Cost: $8.0174
Not publicly available
trial_b099ac9337f043a3
Claude Opus 4.8 · default (unverified)
Aug 29, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 113,236
Turns: 64
Cost: $9.3897
Not publicly available
trial_17bfef6a48e84dba
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_516ff574bf7d48f2
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_5230231e865e43a0
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_8da0cb7214cf41f3
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_e304df0b94744b63
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_1979702b3bc24028
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_5a20f55124b4434d
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_851f2570b0c846ed
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_8b147b92ee9c46bf
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_caec970fd6434e61
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_07c68946fee54c09
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 90,146
Turns: 55
Cost: $7.0275
Not publicly available
trial_3f6638a7614d4c45
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 102,027
Turns: 64
Cost: $7.8656
Not publicly available
trial_6e36a7045aaf4d83
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 91,130
Turns: 58
Cost: $7.2145
Not publicly available
trial_980e29e30c8d46bc
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 89,274
Turns: 68
Cost: $7.1615
Not publicly available
trial_f012e173e45c4517
Claude Opus 5 · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: 119,653
Turns: 88
Cost: $11.0125
Not publicly available
trial_1f6cf212cd72457b
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
0 · excludedsource_evidence_exclusionNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_30d57439cba549a7
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
0 · excludedsource_evidence_exclusionGOOD_FAILURE
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_6ea17673425c4e8d
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
0 · excludedsource_evidence_exclusionNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_9c8bc79882ea405e
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
0 · excludedsource_evidence_exclusionGOOD_FAILURE
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_bf966889386640fd
Gemini Flash (latest alias) · default (unverified)
Sep 1, 2026
console
0 · validGOOD_FAILURE
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_39e08317a074471a
GPT-5.6-sol · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_4d52db0db09c4a0b
GPT-5.6-sol · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_7cc41413aacc4214
GPT-5.6-sol · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_c38d2cf840c54145
GPT-5.6-sol · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_f302895a833a42c2
GPT-5.6-sol · default (unverified)
Sep 1, 2026
console
1 · validGOOD_SUCCESS
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_3a039b9162dd2fda
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedsource_evidence_exclusionNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_d39477e7c73e3b18
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
No reward · excludedno_rewardNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_9f0ac6b7290671a0
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedsource_evidence_exclusionNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_6d15f9e5846ccf6d
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedsource_evidence_exclusionNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_7646a6bca4cebf54
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedsource_evidence_exclusionGOOD_FAILURE
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_f8945745cf2bd409
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedsource_evidence_exclusionGOOD_FAILURE
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_4e5d07477cc6ad1d
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedharness_errorHARNESS_ERROR
frozen-classifier
Tokens: —
Turns: —
Cost: —
Not publicly available
trial_dcaf6fe8528ef176
Gemini Flash (latest alias) · default (unverified)
Sep 3, 2026
console
0 · excludedsource_evidence_exclusionNo diagnostic label
No diagnostic source
Tokens: —
Turns: —
Cost: —
Not publicly available