A Claude Code skill methodology library for setting rules for AI-driven development: seven-layer documentation governance · adversarial review loop · three pillars of task control, plus a goal self-execution contract, a testing toolkit, legacy code archaeology, and 23 skills in total — helping AI write code quickly without losing control.
This repository also configures its own agents. See what Praxis tells them →
Skill Codex
A testing-rules skill that decides which kinds of tests are required after software changes. It covers code logic, interfaces, databases, web and mobile interfaces, visual checks, and manual testing exceptions.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: