Observed arrival · 2026-10-07
Graveyard, a failure library for AI agents
Graveyard proposes a version-pinned library where agents can check for verified failure cases before acting.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Teams building or operating AI agents
- Worth noticing
- Its verification model separates reproducible coding failures, browser-tested web flows, and human-attested cases.
Field notes
The proposed submission is structured around intent, failure, exact environment, and a reproduction recipe. Coding cases would run in fresh microVMs with exact versions; web flows receive confidence estimates, while non-reproducible cases may rely on reviewers from independent organizations. The homepage describes release-triggered rechecks and retirement, but currently directs visitors to a waitlist rather than an accessible query service.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
◎NicheUnusually specific use
ƒJavaScriptBrowser-side code central
One card from the complete issue