Pure-prompt skill that scores how authentically text reads as a real human author's work and flags AI-generated, AI-templated, or derivative spans. Diagnostic, read-only counterpart to humanizer; never rewrites. Works with Claude Code, Cursor, Codex, Antigravity, Gemini, Pi Coder, OpenCode, Copilot, Windsurf, Cline, Continue, Zed, and Aider.
Latest release v1.2.3 — v1.2.3: Drift check CI · 8 Oct 2026
4 files for GitHub Copilot, Codex, OpenCode, Gemini CLI and Windsurf: authenticity-check copilot-instructions.md, authenticity-check AGENTS.md, authenticity-check GEMINI.md, windsurfrules — 2,789 tokens loaded in every session.
.github/copilot-instructions.md A 610 tok AGENTS.md A 944 tok GEMINI.md A 608 tok .windsurfrules A 627 tok These files are hannsxpeter/authenticity-check's own configuration — they tell GitHub Copilot, Codex, OpenCode, Gemini CLI and Windsurf how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.