Structured agent-driven development with multi-judge evaluation — runs a single solution in an isolated forked context with self-critique, and scores it via parallel isolated-context judges. EXECUTE mode (implement + verify) and JUDGE mode (multi-judge debate). One specialist subagent (`sadd-judge` for rubric-based candidate scoring).
Installation
1
Add the marketplace
/plugin marketplace add Git-Fg/taches-principled
2
Install plugins
/plugin
Run these commands in Claude Code to add this plugin to your environment. The marketplace must be added before you can install its plugins.