Mythic

One-liner: Automated LLM vulnerability scanning + RAG evaluation + EU-AI-Act-minded testing

One-liner: Tabular ML validation suites (drift, leakage) — DORMANT since 2024-12

One-liner: YAML-configured CI evals + red-teaming/vulnerability scanning for prompts and agents

One-liner: The RAG evaluation library: faithfulness/relevance/context metrics + testset generation

One-liner: OTel span-level tracing with agentic feedback functions (Snowflake-backed)

One-liner: Runtime input/output validators with self-healing re-asks and structured output

One-liner: UK AISI research-grade eval framework with sandboxed tool execution and 200+ evals

One-liner: The batch scheduling standard: DAGs as Python, 150+ providers