Pinned Loading

1. judge-jitter judge-jitter Public

   CLI to detect score instability in LLM-as-a-judge evals.

   Python 4

2. toklint toklint Public

   Catch tokenizer mistakes before they break your fine-tune.

   Python 4

3. flint flint Public

   VS Code-style autocomplete for Zsh. Local-first, no LLM, no telemetry.

   Rust 1 1