What problem does it solve? Long coding sessions quietly turn assumptions into facts, and confident answers are not always correct ones. This Skill forces every non-trivial decision through an adversarial fresh-context review while course-correction is still cheap, catching hidden failure modes before they reach production. ## Core Features & Use Cases - Structured Doubt Cycle: A five-step loop (CLAIM, EXTRACT, DOUBT, RECONCILE, STOP) that names each decision, isolates the smallest reviewable artifact, and bounds the review to three cycles. - Adversarial Reviewer Prompting: Spawns a fresh-context reviewer biased to disprove, passing only the artifact and contract so the reviewer is never anchored on your conclusions. - Cross-Model Escalation: Optionally routes the artifact through an external CLI (Gemini, Codex) in a read-only sandbox for a second opinion from a different model architecture. - Use Case: Before committing a thread-safety change to a caching layer, you state the claim, hand the diff and its contract to an adversarial reviewer, classify the findings, and fix real issues before the code lands. ## Quick Start Ask the agent to apply doubt-driven development to stress-test the architectural decision or code change you are about to commit.
