Pinned Loading

1. epistemic-skills epistemic-skills Public

   Empirical authoring laws, static linter, and claim-verification harness for AI agent skills (+15.4pp on BenchFlow).

   Python

2. benchmark-designer benchmark-designer Public

   Agent skill for designing and red-teaming AI benchmarks and evals, with a scorer attack battery, judge meta-evaluation, and power analysis.

   Python 2

3. motes motes Public

   Polyphonic granular synthesizer plugin (arm64 AU/VST3). Zero-dependency C DSP core with JUCE editor.

   C++ 3

4. efficient-domain-adaptation efficient-domain-adaptation Public

   Small-data LoRA domain adaptation (Brie) across four base models with multi-judge blind evaluation. DOI: 10.5281/zenodo.17657737.

   Python