Pinned Loading
-
assurancebench
assurancebench PublicA benchmark for evaluating language models on US GAAP and PCAOB-based external audit, with a safety/guardrail suite.
Python
-
auditlm
auditlm PublicOpen, locally-runnable language model and benchmark for US GAAP and PCAOB-based external audit.
Python
-
auditweave
auditweave PublicA tamper-evident audit trail for RAG pipelines and data workflows — trace any AI conclusion back to its evidence.
Python
-
dframe-trace
dframe-trace PublicFind where your pandas/polars pipeline silently broke — which step introduced nulls, dropped rows, or changed a dtype. No rules to write.
Python
-
regradar-backend
regradar-backend PublicAI-powered search and alerts for the US Federal Register
Python
-
unlimited-ocr-eval-harness
unlimited-ocr-eval-harness PublicDeterministic synthetic-OCR corpus + CER/WER eval harness with loop diagnostics and exact ground truth, for measuring quantization cost (MLX).
Python 2
If the problem persists, check the GitHub status page or contact support.