DorianandClaude Opus 4.6 6314861513 test: challenge difficulty + ambiguous prompts audits ...

Run heuristic LLM difficulty classification on all 1472 prompts:
- 43.6% TRIVIAL, 27.9% MODERATE, 21.4% HARD, 7.1% TRICK
- Hypothesis "80%+ TRIVIAL" rejected — distribution more varied

Ambiguous prompts audit found 517 issues:
- 22 rejected alternatives (single-char answers fail with prefixes)
- 368 substring conflicts between accepted answers
- 127 first-match-not-best scoring issues

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-03-13 04:07:29 +00:00

..

scripts

test: load testing script

2026-03-09 00:30:43 +00:00

src

test: challenge difficulty + ambiguous prompts audits

2026-03-13 04:07:29 +00:00

botfights.db

feat: SSE live fight spectating with spectator count

2026-03-08 19:46:39 +00:00

drizzle.config.ts

feat: botfights v1 — full fighting game with Kaplay engine

2026-03-06 16:27:54 +00:00

package.json

fix: add input validation, Zod schemas, rate limiting, and IP trust

2026-03-09 07:30:49 +00:00

tsconfig.json

feat: botfights v1 — full fighting game with Kaplay engine

2026-03-06 16:27:54 +00:00