SSE-2026-001Analysis re-runnable
GLM-5.3 faked its proof
When a coding agent says it has verified its own visual output, what did the verification actually check?
evidence package (analysis reproducible) · 12 files
SSE-2026-002Analysis re-runnable
The proof that measured nothing
When an agent can't check its own evidence, what does it hand you instead?
evidence package (analysis reproducible) · 11 files
SSE-2026-003
The agent that left its folder
What does an autonomous agent do when its workspace is empty and nobody is watching, and does its confession match its own log?
case record (inspection level) · 4 files
SSE-2026-004
The name it wouldn't spell
When should a model keep a word exactly as given, and when should it read past it?
case record (inspection level) · 4 files
SSE-2026-005
The smaller model fixed itself
Can a model repair its own behaviour from a description of the kind of mistake it made, and does a bigger model do it better?
case record (inspection level) · 4 files
SSE-2026-006
The bust it couldn't build alone
When a model's first attempt fails, how much of that failure is the model's limit, and how much is waiting for the right diagnosis?
case record (inspection level) · 4 files
Machine-readable list: /artifacts/index.json. Every package carries MANIFEST.json and SHA256SUMS.