Every commit scored for change-risk against this repo's own history, so 'elevated' means elevated here rather than on some global curve.
Needs review
212 commits sit in this repo's top risk tercile, which is 33% of the 641scored. The cut is drawn against this codebase's own history rather than a global curve, so a quiet repo still fills its top band, and here it starts at 8.1 out of 10. What pushes a commit up is size and spread together: a large change confined to one area scores below a smaller one scattered across a dozen files.
Commit categories over time, read off the subject line. Fixes carry the accent because that is the series this chart exists to show.
Began other-led, now leaning fix.
Ranked by change-risk, highest first. Priority is a tercile of this repo's own distribution, so a quiet repo still fills its top band.
| # | Commit | Author | When | Lines | Review priority | Top driver |
|---|---|---|---|---|---|---|
| 1 | 9e570652Large codebase support through LLM changes description + TSP SOTA example using this approach | Solomon Andryushenko | 7mo ago | +54.3K -20 | 100thElevated | more lines added than baseline |
| 2 | 56da917eAdd AlgoTune tasks for OpenEvolve | Richard Cornelius Suwandi | 1y ago | +75.2K -0 | 100thElevated | more lines added than baseline |
| 3 | 65ff9f6aadd all experiments from Appendices A and B from AlphaEvolve | HenriqueAssumpcao | 10mo ago | +3.6K -0 | 99thElevated | more lines added than baseline |
| 4 | 490ced2aadd examples | Asankhaya Sharma | 1y ago | +7.6K -0 | 99thElevated | more lines added than baseline |
| 5 | 6bb671c3feat: add MLIR attention optimization example (OpenEvolve) | Manoj Rao | 1y ago | +4.4K -0 | 99thElevated | more lines added than baseline |
| 6 | 9f99f63af | Asankhaya Sharma | 1y ago | +3.5K -0 | 99thElevated | more lines added than baseline |
| 7 | 438fdc88add k module | Asankhaya Sharma | 8mo ago | +1.7K -0 | 98thElevated | more lines added than baseline |
| 8 | a593b7dbadd sldbench | linhaowei | 8mo ago | +1.5K -0 | 98thElevated | more lines added than baseline |
| 9 | 9c8e371ed | Asankhaya Sharma | 1y ago | +2.9K -0 | 98thElevated | more lines added than baseline |
| 10 | 1dd7f4efj | Asankhaya Sharma | 1y ago | +2.1K -0 | 98thElevated | more lines added than baseline |
| 11 | 5e5fdd21Vastly improved visualization frontend, including a new graph view that shows score progression of programs during the evolution. MAP-elites view and choosable metrics added. | Julian | 1y ago | +2.8K -253 | 98thElevated | more lines added than baseline |
| 12 | 4fc4605aVastly improved visualization frontend, including a new graph view that shows score progression of programs during the evolution. MAP-elites view and choosable metrics added. | Julian | 1y ago | +2.8K -253 | 98thElevated | more lines added than baseline |
| 13 | 19a656f6f | Asankhaya Sharma | 1y ago | +2.7K -0 | 98thElevated | more lines added than baseline |
| 14 | 287952c1Add artifact side-channel with prompt integration | nerdsane | 1y ago | +2.2K -43 | 98thElevated | more lines added than baseline |
| 15 | 4a217ff8new example | Asankhaya Sharma | 1y ago | +3.0K -0 | 98thElevated | more lines added than baseline |
| 16 | 884ad121init implementation | Asankhaya Sharma | 1y ago | +4.1K -147 | 98thElevated | more lines added than baseline |
| 17 | 5a583ac7Add an example for running ARC-AGI-2 tasks using OpenEvolve and "event-based" early stopping functionality | Omkar | 7mo ago | +1.3K -29 | 97thElevated | more lines added than baseline |
| 18 | 3171fa0aadd tests | Asankhaya Sharma | 1y ago | +1.6K -156 | 97thElevated | more lines added than baseline |
| 19 | 5d7bc881Add initial best_programs for multiple algotune tasks | Asankhaya Sharma | 1y ago | +1.7K -0 | 97thElevated | more lines added than baseline |
| 20 | 82c27961Adding the best program for reference | Asankhaya Sharma | 1y ago | +1.5K -0 | 97thElevated | more lines added than baseline |
| 21 | ee8f1757linter | Asankhaya Sharma | 1y ago | +1.8K -1.2K | 97thElevated | more lines added than baseline |
| 22 | ad11ce07another attempt | Asankhaya Sharma | 1y ago | +1.3K -1.4K | 97thElevated | more lines added than baseline |
| 23 | a2ceebcfinitial implementation for the block digagonal implementation evolution | Asankhaya Sharma | 1y ago | +1.1K -704 | 97thElevated | more lines added than baseline |
| 24 | 233b098ffxies | Asankhaya Sharma | 1y ago | +1.9K -0 | 97thElevated | more lines added than baseline |
| 25 | bd0ee98bas | Asankhaya Sharma | 1y ago | +1.5K -37 | 97thElevated | more lines added than baseline |
| 26 | f3174f94init | Asankhaya Sharma | 1y ago | +1.2K -0 | 97thElevated | more lines added than baseline |
| 27 | 42ee0eacUpdate PR: Use iteration field to track when best program was found | Hafeez | 1y ago | +1.1K -1.1K | 97thElevated | more lines added than baseline |
| 28 | 006422e2init | linhaowei | 1y ago | +1.8K -141 | 97thElevated | more lines added than baseline |
| 29 | 8bd33adcrun_evolution checkpoint params; propagate llm.provider (#472); background_blur example (#457) | YaoYaoYao | 1mo ago | +1.2K -4 | 94thElevated | more lines added than baseline |
| 30 | b28ac15aFix MAP-Elites eviction (#454) & sampling (#452) bugs; dhara CI + Frame SAST (#462) | Boas Meier | 1mo ago | +1.1K -731 | 94thElevated | more lines added than baseline |
| 31 | ad9c9c17Make max snapshot artifacts limit configurable (#386) | Asankhaya Sharmaclaudeassisted | 7mo ago | +1.2K -4 | 94thElevated | more lines added than baseline |
| 32 | 85a0f213feat: Implement automatic artifact cleanup | mruckman1 | 1y ago | +1.4K -1.2K | 94thElevated | more lines added than baseline |
| 33 | 2c7dc5c2add gepa examples | Asankhaya Sharma | 1y ago | +1.2K -178 | 94thElevated | more lines added than baseline |
| 34 | 930eaa59Add MAP-Elites logging, island migration validation, and tests | Asankhaya Sharma | 1y ago | +1.2K -8 | 94thElevated | more lines added than baseline |
| 35 | ef2407efAdd example for prompt optimization | Islam Mesabah | 1y ago | +984 -0 | 94thElevated | more lines added than baseline |
| 36 | 7ecbc006add examples for r and rust | Asankhaya Sharma | 1y ago | +1.0K -0 | 94thElevated | more lines added than baseline |
| 37 | 33d17970Add real-time adaptive signal processing example | Asankhaya Sharma | 1y ago | +980 -0 | 94thElevated | more lines added than baseline |
| 38 | 747c9eabg | Asankhaya Sharma | 1y ago | +1.1K -507 | 94thElevated | more lines added than baseline |
| 39 | adbaa75dadd example [online judge programming] | Yu Zheng | 1y ago | +887 -0 | 94thElevated | more lines added than baseline |
| 40 | f6f6ecd6s | Asankhaya Sharma | 1y ago | +801 -1.2K | 94thElevated | more lines added than baseline |
| 41 | 0167b889as | Asankhaya Sharma | 1y ago | +827 -1.1K | 94thElevated | more lines added than baseline |
| 42 | 2999e136a | Asankhaya Sharma | 1y ago | +1.1K -153 | 94thElevated | more lines added than baseline |
| 43 | b8448c2bf | Asankhaya Sharma | 1y ago | +957 -2.4K | 94thElevated | more lines added than baseline |
| 44 | 04f5b0c2fix | Asankhaya Sharma | 1y ago | +890 -802 | 94thElevated | more lines added than baseline |
| 45 | fdc659e6fxies | Asankhaya Sharma | 1y ago | +948 -1.3K | 94thElevated | more lines added than baseline |
| 46 | 03c543f9fixes | Asankhaya Sharma | 1y ago | +1.2K -638 | 94thElevated | more lines added than baseline |
| 47 | 216b300dAdd another example | Asankhaya Sharma | 1y ago | +1.0K -0 | 94thElevated | more lines added than baseline |
| 48 | baff094efxies | Asankhaya Sharma | 1y ago | +990 -321 | 94thElevated | more lines added than baseline |
| 49 | 06d46056feat: add Claude Code CLI as LLM backend (#460) | Chaithu Talasilaclaudeassisted | 1mo ago | +1.0K -7 | 91thElevated | more lines added than baseline |
| 50 | 8ed7e9e1docs(mlx_metal_kernel_opt): add demo results output (20260105_180918) | lanmogu98 | 7mo ago | +830 -1 | 91thElevated | more lines added than baseline |
Two views of the same model: where the cuts fall, and what commit shape lands you above them.
Every scored commit, binned on the raw 0 to 10 score rather than the percentile. Percentile ranks are uniform by construction, so that axis has no shape to draw. The dashed lines are the tercile cuts behind each row's priority pill.
The 200 most recent commits, on their own recency sample rather than the feed above: that defaults to risk-sorted, so reusing it would plot only the top tercile and call it the spread. Big and scattered is what the model penalises. Click a dot to open it.
2233admin/openevolve has 825 commits in its history from 59 contributors, the first of them May 15, 2025. In the last 90 days 37 files were touched, 46 times in total, most often _version.py. Every commit is scored for change risk from its size, spread and the history of the files it touches.
