9 projects
rce-path
Local-first, evidence-backed Python execution-path investigation
promptsanitizer
Secrets firewall for AI pipelines — redact credentials, PII, prompt injection, and RCE payloads before they reach LLMs
cv-quality
Dataset-centric CV toolkit: label-quality checks, mislabel detection, class-imbalance analysis, and active-learning loop orchestration.
rai-check-ml
Fairness, robustness, and data quality audits for traditional ML models
rai-check-genai
Generative AI audits for LLM, RAG, and agentic AI systems
rai-check-dl
Deep learning audits for image classification, medical imaging, and scientific AI
rai-check-core
Shared audit engine, findings, reports, and CLI for the RAI Check Kit
rai-check-kit
Responsible AI (RAI) Check Kit - security, reliability, and trustworthy AI system checks
agentdog
Lightweight evaluation toolkit for AI agents — test tool use, grounding, safety, and efficiency before production
