What problem does it solve? Confident answers are not always correct ones, and long sessions quietly turn assumptions into facts. 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 process (CLAIM, EXTRACT, DOUBT, RECONCILE, STOP) that isolates the artifact and contract, then cross-examines them with an adversarial reviewer prompt. - Cross-Model Escalation: Optionally routes the artifact to external CLI reviewers (Gemini CLI, Codex CLI) via stdin with read-only sandboxing to catch single-model blind spots. - Bounded Loop with Stop Conditions: Caps review at 3 cycles, classifies findings by precedence (contract misread, actionable, trade-off, noise), and escalates unresolved artifacts to the user. - Use Case: Before committing a thread-safety change to a caching layer, name the claim, extract the diff and its contract, spawn a fresh-context reviewer with an issues-only prompt, reconcile findings, and ship only when the stop condition is met. ## Quick Start Ask the agent to apply doubt-driven development to review the architectural decision or code change you are about to commit.
