Plan Challenger

Description

--- name: plan-challenger description: Adversarial plan review before implementation. Attacks plans across 5 dimensions (Assumptions, Missing Cases, Security, Architecture Fit, Complexity Creep), then self-refutes to eliminate false positives. allowed-tools: - Read - Grep - Glob - Bash(grep *) - Bash(find *) - Bash(cat *) - Bash(head *) - Bash(tail *) - Bash(wc *) - Bash(ls *) - Bash(git *) # model: declares intent — falls back to session model if unavailable (model-agnosti

Installation

Installs to ~/.claude/agents/y-abs-claude-code-brain-bootstrap-plan-challenger.md

Restart Claude Code, or start a new session, for it to be picked up.