Integrity
Coverage
- 603 logical evidence files
- 54 compatible groups
- 30 problems
Subject readiness, tier recall, and exporter problems; missing values never become zero
Answer-key coverage
Every run-referenced subject remains visible, including subjects with no scored trial
| Subject | Tier counts | Defects | Runs | Action |
|---|---|---|---|---|
| feat(proxy): reclaim space from the blob store | T1 3 / T2 7 / T3 6 / T4 4 | 20 | 74 | Inspect subject |
Tier recall matrix
Each row is one cohort, reviewer, configuration, and build; values are exporter medians within that row
| Reviewer | Subject | Configuration | Build | Scored / recorded | Tier 1 | Tier 2 | Tier 3 | Tier 4 |
|---|---|---|---|---|---|---|---|---|
| afi afi · cli | proxy | default Default profile · OpenAI · gpt 5.6 sol high | afi 0.30.0 | 3/3 | 100% | 42.9% | 50% | 50% |
| afi afi · cli | proxy | default Default profile · xAI · grok 4.5 high | afi 0.30.0 | 2/2 | 66.7% | 35.7% | 50% | 25% |
| afi afi · cli | proxy | default Default profile · Z.AI · glm 5.2 high | afi 0.30.0 | 1/1 | 0% | 28.6% | 33.3% | 0% |
| afi afi · cli | proxy | default Default profile · Meta · llama 4 maverick high | afi 0.30.0 | 0/3 | Not recorded | Not recorded | Not recorded | Not recorded |
| afi afi · cli | proxy | default Default profile · Z.AI · glm 5.3 flash high | afi 0.30.0 | 1/1 | 33.3% | 42.9% | 50% | 25% |
| afi afi · cli | proxy | default Default profile · Mistral AI · devstral 2512 high | afi 0.30.0 | 0/1 | Not recorded | Not recorded | Not recorded | Not recorded |
| afi afi · cli | proxy | default Default profile · Minimax · m3 high | afi 0.30.0 | 1/1 | 33.3% | 0% | 16.7% | 25% |
| afi afi · cli | proxy | default Default profile · Moonshot AI · kimi k2.7 code high | afi 0.30.0 | 1/1 | 33.3% | 28.6% | 33.3% | 25% |
| afi afi · cli | proxy | default Default profile · DeepSeek · v4 flash high | afi 0.30.0 | 1/1 | 0% | 0% | 0% | 0% |
| afi afi · cli | proxy | default Default profile · Z.AI · glm 5.3 high | afi 0.30.0 | 1/1 | 33.3% | 57.1% | 33.3% | 25% |
| afi afi · cli | proxy | default Default profile · DeepSeek · v4 pro high | afi 0.30.0 | 1/1 | 66.7% | 57.1% | 66.7% | 50% |
| afi afi · cli | proxy | default Default profile · xAI · grok 4.5 high | afi 0.30.0 | 1/1 | 66.7% | 28.6% | 50% | 25% |
All 54 comparison groups
proxy
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
- default afi
Data quality
Paid, malformed, and partial records remain visible · showing 30 of 30 across 5 pages
afi wrote no findings document
afi/devstral-2512/20260831T163155Zafi wrote no findings document
afi/devstral-2512/20260831T163156Zafi wrote no findings document
afi/devstral-2512/20260831T163157Zafi wrote no findings document
afi/llama-4-maverick/20260831T163158Zafi wrote no findings document
afi/llama-4-maverick/20260831T163159Zafi wrote no findings document
afi/llama-4-maverick/20260831T163200Zafi wrote no findings document
afi/opus-5/20260831T152641Zafi wrote no findings document
afi/opus-5/20260831T154400Zafi wrote no findings document
afi/qwen3-coder-30b/20260831T111207Zafi wrote no findings document
afi/qwen3-coder-30b/20260831T111208Zafi wrote no findings document
afi/qwen3-coder-30b/20260831T111209Zafi wrote no findings document
afi/qwen3.7-flash/20260831T125923Zafi wrote no findings document
afi/qwen3.7-flash/20260831T125924Zafi wrote no findings document
afi/qwen3.7-flash/20260831T125925Zunscored trial excluded from aggregate metrics
afi/llama-4-maverick/20260831T163158Zunscored trial excluded from aggregate metrics
afi/llama-4-maverick/20260831T163159Zunscored trial excluded from aggregate metrics
afi/llama-4-maverick/20260831T163200Zunscored trial excluded from aggregate metrics
afi/devstral-2512/20260831T163155Zunscored trial excluded from aggregate metrics
afi/devstral-2512/20260831T163156Zunscored trial excluded from aggregate metrics
afi/devstral-2512/20260831T163157Zunscored trial excluded from aggregate metrics
afi/opus-5/20260831T152641Zunscored trial excluded from aggregate metrics
afi/qwen3-coder-30b/20260831T111207Zunscored trial excluded from aggregate metrics
afi/qwen3-coder-30b/20260831T111208Zunscored trial excluded from aggregate metrics
afi/qwen3-coder-30b/20260831T111209Zunscored trial excluded from aggregate metrics
afi/qwen3.7-flash/20260831T125923Zunscored trial excluded from aggregate metrics
afi/qwen3.7-flash/20260831T125924Zunscored trial excluded from aggregate metrics
afi/qwen3.7-flash/20260831T125925Zunscored trial excluded from aggregate metrics
afi/opus-5/20260831T154400Zsession records 'failed', but the linked run lifecycle records 'completed'
afi/deepseek-v4-pro/20260831T111226Zplanned run has no native record yet
afi/sonnet-5/20260831T144454Z