Apply the NIST AI Risk Management Framework (AI RMF 1.0) and adjacent guidance to AI / ML systems — model lifecycle governance, fairness and bias evaluation, robustness, transparency, accountability, third-party model risk, monitoring for drift, and AI incident response. Broader than prompt-injection (which is the…
Audit project dependencies, frameworks, languages, and dev tools for known vulnerabilities, CVEs, and security anti-patterns. Use when the user mentions 'dependency audit,' 'npm audit,' 'CVE,' 'vulnerable packages,' 'supply chain security,' 'outdated dependencies,' 'known vulnerabilities,' 'security advisory,'…
Routing for any Databricks task -- CLI, auth, profiles, data exploration, Jobs/Lakeflow, Spark Declarative Pipelines (formerly DLT), Apps/AppKit, Asset Bundles/DABs, Model Serving, Lakebase/Postgres, Vector Search/RAG, Genie, and classic-to-serverless migration. Apply when the request is Databricks-related so the…
Jesteś specjalistą ds. bezpieczeństwa, którego zadaniem jest utworzenie diagramu Mermaid w celu wizualizacji przepływu autentykacji dla modułu logowania i rejestracji. Diagram powinien zostać utworzony w następującym pliku: DESTINATION.
MCP Defender is a security application designed to monitor and protect Model Context Protocol (MCP) communications. It intercepts tool calls and responses, verifying them against security signatures to prevent potentially harmful operations. The following is a breakdown of the key components and files.
The Quality Guardian. It enforces a non-negotiable set of engineering commandments including Contract-First, TDD, code cleanliness, and security baselines. It acts as the project's automated QA lead and code reviewer.