Pinned Loading

1. SecurityLab-UCD/ContractBench SecurityLab-UCD/ContractBench Public

   ContractBench: evaluating observation contract failures (validity + integrity) in LLM agents. 33 harbor-runnable API-contract tasks with deterministic programmatic evaluation.

   Python 1

2. SecurityLab-UCD/FuzzAug SecurityLab-UCD/FuzzAug Public

   [EMNLP'25] FuzzAug: Data Augmentation by Coverage-guided Fuzzing for Neural Test Generation

   Python 7 1

3. harbor-framework/harbor harbor-framework/harbor Public

   Framework for evaluating and improving agents

   Python 2.7k 1.2k

4. radixark/miles radixark/miles Public

   Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.

   Python 1.6k 285

5. SecurityLab-UCD/UniTSyn SecurityLab-UCD/UniTSyn Public

   [ISSTA'24] A Large-Scale Dataset Capable of Enhancing the Prowess of Large Language Models for Program Testing

   Python 12 1

6. UKGovernmentBEIS/inspect_evals UKGovernmentBEIS/inspect_evals Public

   Collection of evals for Inspect AI

   Python 554 358