kayneai/code-auditor

This is a zero-stars, zero-forks repository with no development velocity over 76 days—strong signals of an abandoned personal experiment. The core capability (AI-powered code auditing via LLM) is a thin wrapper around frontier lab APIs (OpenAI, Anthropic, etc.) applied to GitHub repo scanning. No evidence of novel architecture, custom models, domain-specific training, or novel analytical approach—just LLM + GitHub API orchestration. The problem space (code quality/security scanning) is actively competed by: (1) GitHub's native code scanning and Dependabot, (2) Snyk, Semgrep, SonarQube, and other established SAST tools, (3) frontier labs themselves (Copilot for code review, Claude for code analysis). Frontier risk is HIGH because: OpenAI and Anthropic could trivially add 'audit my GitHub repo' as a native feature in their web UI or API ecosystem. The tool offers no switching costs, no data gravity, no irreplaceable dataset or model—it's a consumer application built on their inference. Complete lack of adoption, maintenance, or community indicates the builder has moved on or found the approach non-viable. Easy to clone, trivially reproducible, no moat.