Python CLI and agent plugin that audits whether an AI agent does the job it is supposed to do, using adversarial and operational tests.
Builds a queryable knowledge graph from a codebase and its docs, PDFs, and configs, driven by a /graphify skill for AI coding assistants.
Local CLI, MCP server, and catalog for searching, selecting, validating, and previewing installs of 2,700+ reusable SKILL.md agent playbooks.