Auto-Arch Tournament vs FoxGuard

Auto-Arch Tournament is an autonomous research system where an AI agent iteratively proposes, implements, and validates microarchitectural improvements to a RISC-V CPU. Starting from a standard 5-stage pipeline, the loop runs hypotheses in parallel, each going through formal verification (53 symbolic checks), cycle-accurate simulation, multi-seed FPGA place-and-route, and CoreMark CRC validation. Only hypotheses that beat the current champion get merged; everything else gets discarded. Starting from 301 iterations/second, the system hit 577 iter/s (+92%) across 73 attempts in 9.8 hours — producing a design 26% faster and 40% smaller in LUTs than the baseline. The insight the author drives home is that the real innovation isn't the AI agent — it's the verifier. The orchestrator is hardcoded to prevent agents from manipulating their own evaluation gates, a simple but critical design constraint that turns a creative process into a trustworthy one. Without a rigorous verification harness, agent-driven optimization becomes a confidence trick. This is early but fascinating proof that AI-driven hardware design loops can produce commercially meaningful gains. The repo uses Claude Code or Codex as the coding agent, SystemVerilog for the RTL, and standard open-source EDA tooling (Yosys, nextpnr, Verilator). It's a compelling template for anyone building agentic optimization loops where correctness matters.

FoxGuard is a Rust-based security scanner designed to run at linter speed — sub-second full-project scans with zero cold-start overhead. Built on tree-sitter for real AST parsing (not regex heuristics), it covers 100+ security rules across 10 languages including Python, JavaScript, TypeScript, Go, Java, and Rust. Rules cover SQL injection, XSS, command injection, path traversal, hardcoded credentials, insecure deserialization, and more. Ships as a single native binary with no JVM or Python runtime dependency. FoxGuard is explicitly designed for the pre-commit and CI hook workflow that AI-generated code has made more important. With agents writing hundreds of lines per session, manual code review is increasingly the bottleneck — FoxGuard runs in the background on every save or commit and surfaces security anti-patterns before they hit a PR. The rule set is MIT-licensed and community-extensible via YAML definitions. For teams using AI coding agents, the "AI writes fast, security doesn't keep up" gap is real. FoxGuard positions itself as the fast-path answer: not a full SAST platform, but a zero-friction first-pass filter that catches the obvious issues before they accumulate into an audit finding.