More technical detail (so this thread doesn’t become “sounds cool”):
- Data format: each evaluation is structured JSON in /data, and the site renders from that.
- Scoring: overall score is computed from the 5 dimensions; you can compute custom scores by applying your own weights (CVSS-style).
- Evidence: each dimension includes citations + a confidence level, so disagreements can be about the evidence, not vibes.
If you want to critique one thing: pick a single evaluation you know well and tell me what’s wrong/missing in the evidence, then we’ll fix the methodology or the rubric.
More technical detail (so this thread doesn’t become “sounds cool”):
- Data format: each evaluation is structured JSON in /data, and the site renders from that. - Scoring: overall score is computed from the 5 dimensions; you can compute custom scores by applying your own weights (CVSS-style). - Evidence: each dimension includes citations + a confidence level, so disagreements can be about the evidence, not vibes.
If you want to critique one thing: pick a single evaluation you know well and tell me what’s wrong/missing in the evidence, then we’ll fix the methodology or the rubric.