Find code blind spots with independent AI reviews.
Synthesizes independent adversarial reviews of code changes into a verdict with prioritized findings.
When to use it
Use when asked for an adversarial or multi-model review, to stress-test code, or to find blind spots.
Give it code changes to review; it returns a synthesized verdict without automatically applying changes.
What you provide
Uses its Subagents guidance to choose each reviewer's reasoning level and speed; whether installation is required is unstated.
This skill
interrogate default model table
Proposes reviewer model default updates
~/.cursor/rules/pstack-models.mdc
Reads Cursor reviewer model settings
~/.claude/rules/pstack-models.md
Reads Claude Code reviewer settings