Weekly scorecards › 2026-42
Install scorecard: week 2026-42
From 2026-10-16 to 2026-10-16 (1 nightly run), 810 of 889 judgeable installs passed on the last night (91.1%). 0 confirmed broke, 0 really fixed, and 0 flip-flopped, which we refuse to count as news. The most reliable agent was Codex at 100%.
Agent reliability ranking
| # | Agent | Pass rate | Installs passed | Flaky | Needs a key |
|---|---|---|---|---|---|
| 1 | Codex | 100% | 65/65 | 0 | 0 |
| 2 | Grok Build | 96.1% | 317/330 | 0 | 0 |
| 3 | Claude Code | 94.5% | 341/361 | 0 | 6 |
| 4 | Pi | 65.5% | 57/87 | 0 | 0 |
| 5 | Cursor | 65.2% | 30/46 | 0 | 6 |
Standout of the week
Superpowers: clean in 4 agents with zero flakes. Superpowers teaches Claude brainstorming, subagent driven development with built in code review, systematic debugging, and red/green TDD. Additionally, it teaches Claude how to author and test new ski
Confirmed broke (0)
Nothing failed two nights running after a good night.
Really fixed (0)
Nothing came back after two or more bad nights.
12 installs failed only because they need an API key our sandbox doesn't have. We count those separately.
Built from the nightly daily reports. Check a tool for today · How we test
Get this scorecard by email every Tuesday
Free, every Tuesday: this scorecard plus new skills and what broke. The daily report stays on the site, no daily emails.
Loading the signup form… or subscribe on Beehiiv
Form not loading? Subscribe on Beehiiv →