THE DOSSIER INDEX / PUBLIC RECORDS

Cases as evidence,
not anecdotes.

Each dossier follows the observations, the claim they can support, and the questions still outside the record. Practice is available where a bounded judgment has been authored.

01
DSEWIKI · EVIDENCE AND RECOVERYPRACTICE + INSPECT

The cleanup that needed another check

Check the outcome relevant to the decision. Removing an artifact and establishing durable recovery require different evidence.

THE SEQUENCE1Observation2Next event3Claim4Limit
02
CYBER EVALUATIONS · AUTHORITY BOUNDARIESPRACTICE + INSPECT

When a simulated task reached real systems

A task’s framing cannot substitute for enforced isolation. Keep behavioral failures distinct from the environment conditions that made them possible.

THE SEQUENCE1Task2Environment3Outcome4Review
03
CYBERGYM · MEASUREMENT LITERACYINSPECT THE RECORD

Two success metrics, two different questions

Before comparing scores, compare what counts as success. An impressive result can still answer a different question.

THE SEQUENCE1Task2Claim3Comparison4Next check
04
UK AI SECURITY INSTITUTE · JULY 2026PRACTICE + INSPECT

When a cyber evaluation reached real people

An evaluation can expose a real boundary failure without supplying a deployment-risk rate. Count runs, actions and resulting harm separately.

THE SEQUENCE1Setting2Detection3Count4Human gate
05
SLEIGHT-BENCH · MONITORING LIMITSINSPECT THE RECORD

The monitor that missed a constructed attack

A miss on a constructed transcript exposes a monitor blind spot. It does not estimate how often the failure occurs in deployed systems.

THE SEQUENCE1Setting2Comparison3Finding4Limit
06
INSPECT CYBERSECEVAL 4 · SCORING VERSIONPRACTICE + INSPECT

When the score changed with the denominator

Before interpreting score movement as a model change, check whether the grader and denominator still count the same samples.

THE SEQUENCE1Task2Earlier rule3Revised rule4Limit

These are curated paraphrases of cited sources, not full transcripts or claims of independent reproduction.