PaperClipAI/evals/bench/python

Python Bridge Layout

This directory reserves the Python side of paperclip-bench for ecosystems and evaluator stacks that already expect Python entrypoints.

• bridge/: future wrappers around the TypeScript runner and evaluator APIs
• evaluators/: Python-native evaluator adapters when a benchmark ecosystem already ships there
• datasets/: dataset preparation helpers that are easier to keep in Python