Python CLI and agent plugin that audits whether an AI agent does the job it is supposed to do, using adversarial and operational tests.
CLI and setup toolkit that gives AI agents read and search access to Twitter, Reddit, YouTube, GitHub, Bilibili, and XiaoHongShu.
Scans git repos, files, and stdin for hardcoded secrets such as passwords, API keys, and tokens.
Packages LLMs and a whisper-based speech-to-text tool into single executable files that run locally without installation.