Every run on every task
Language
Kind
Solved Not solvedHover over a run for detailed stats
| Task | Opus 5.5 | Fable 5.2 | GPT-6 Astra | Fable 5.1 | GPT-6.1 Terra | GPT-6.1 Sol | Haiku 5.5 | Sonnet 5.5 | GLM-5.3 Flash | GPT-6 Luna | Solved |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Hundreds of players missing from an importPython · Bug fix | 0 / 19 | ||||||||||
| Pages load with their numbers blank once the season startsTypeScript · Bug fix | 1 / 19 | ||||||||||
| One provider's teams appear with nobody on themTypeScript · Bug fix | 4 / 19 | ||||||||||
| When offers are equal, the sale is credited to the losing buyerPython · Bug fix | 10 / 19 | ||||||||||
| One failed check silences the other alertsPython · Feature | 11 / 19 | ||||||||||
| Choosing "Uncategorized" doesn't stickJavaScript · Bug fix | 12 / 19 | ||||||||||
| Serve a second user from one installPython · Feature | 12 / 19 | ||||||||||
| Derive player values instead of hand-picking two numbersTypeScript · Feature | 12 / 19 | ||||||||||
| Watch a list of events instead of onePython · Feature | 12 / 19 | ||||||||||
| Every player from one provider shows zero pointsTypeScript · Bug fix | 13 / 19 | ||||||||||
| Computer bidders hoard their fundsTypeScript · Bug fix | 13 / 19 | ||||||||||
| Import fails when the week has no data yetPython · Bug fix | 14 / 19 | ||||||||||
| Flag ledger entries that clash with earlier onesPython · Feature | 15 / 19 | ||||||||||
| Some intercepted calls are refused before the service sees themRust · Bug fix | 16 / 19 | ||||||||||
| Build a practice draft with computer opponentsTypeScript · Feature | 16 / 19 | ||||||||||
| An expired session makes a button hangJavaScript · Bug fix | 17 / 19 | ||||||||||
| Private screenshots open for outsiders, public ones don'tJavaScript · Bug fix | 17 / 19 | ||||||||||
| Server takes traffic after a failed startupPython · Bug fix | 17 / 19 | ||||||||||
| Remote files misbehave after a reconnectRust · Bug fix | 18 / 19 | ||||||||||
| A slow page is read as sold out, and no alert firesPython · Bug fix | 18 / 19 | ||||||||||
| A raw file link skips the workspace's privacy rulesJavaScript · Bug fix | 19 / 19 | ||||||||||
| A public workspace's new upload asks for sign-inJavaScript · Bug fix | 19 / 19 | ||||||||||
| Sharing from a phone skips the upload capJavaScript · Bug fix | 19 / 19 | ||||||||||
| One bad record drops the rest of an importPython · Bug fix | 19 / 19 | ||||||||||
| Give every short link an ownerJavaScript · Feature | 19 / 19 | ||||||||||
| Let anyone look up where a short link goesJavaScript · Feature | 19 / 19 | ||||||||||
| Answer a browser extension in JSON, not redirectsJavaScript · Feature | 19 / 19 | ||||||||||
| Paying doesn't upgrade the planJavaScript · Bug fix | 19 / 19 | ||||||||||
| Reject a config that silently drops trafficRust · Bug fix | 19 / 19 | ||||||||||
| Recognize a venue's own waiting roomPython · Feature | 19 / 19 |
Each mark is one run of the task by that agent. An agent with one mark a task has run every task once.