# Viv

> Viv is an AI agent for digital verification that automatically debugs failing test cases in hardware designs.

Viv is developed by [Silogy](https://silogy.io). It analyzes code, logs, waveform files, and specifications to find the root cause of bugs in RTL and testbench code.

## Key capabilities

- Automated root cause analysis of failing test cases in digital verification
- Efficient analysis of multi-gigabyte waveform files (FST, VCD, FSDB)
- Interactive follow-up chat after analysis completes
- Re-run analysis with different hypotheses
- Export bug reports to Markdown
- VIV.md repository guides for faster, more targeted debugging
- Integration with third-party data sources via MCP servers

## Setup

- Runs entirely on-premise — no data leaves your infrastructure
- Installs in seconds via npm or a one-line shell command
- Requires Node.js 20+ and an LLM API key (OpenAI, Claude, Llama, or vLLM-compatible)
- Available as a CLI tool and as a VS Code extension

## Interface

- CLI: `viv submit` to run analysis, `viv chat` for chat-only mode, `viv history` to browse past analyses
- VS Code extension for an integrated UI experience
- Headless mode for CI/CD pipelines (Markdown report to stdout)

## Links

- [Documentation](https://docs.silogy.io/)
- [Setup guide](https://docs.silogy.io/setup)
- [CLI reference](https://docs.silogy.io/cli-reference)
- [Website](https://silogy.io)