Sentinel DV

About

Sentinel DV is an open-source Model Context Protocol (MCP) server designed to bridge the gap between complex verification ecosystems and AI agents. It offers safe, structured, and read-only access to critical verification artifacts from frameworks like UVM, cocotb, and SystemVerilog, including waveform summaries. By providing a unified, schema-driven interface with built-in security features such as automatic redaction and path sandboxing, Sentinel DV enables deterministic triage, root-cause analysis, and deep verification insights without exposing raw logs or granting control over simulators.

Key Features

• Rich access to verification data, including test results, UVM topology, failure analysis, assertions, and coverage metrics.
• Schema-driven contracts providing versioned, JSON Schema-validated APIs for deterministic data access.
• Efficient data indexing with DuckDB and smart pagination for scalable performance and bounded outputs.
• 1 GitHub stars
• Security-first design with read-only access, automatic data redaction, and path sandboxing.
• Simulator-agnostic integration through an adapter pattern, supporting any simulator without runtime dependencies.

Use Cases

• Enable AI agents to perform deterministic triage and root-cause analysis of verification test failures.
• Provide detailed insights into assertion failures and UVM topology within specific verification components.
• Conduct comparative analysis of regression runs and track functional or code coverage across different builds.