What problem does it solve? After code is written, its own tests often confirm the author's assumptions rather than challenge them. This Skill applies independent, adversarial validation that tries to break the implementation, exposing edge cases, security holes, and contradictions before they reach production. ## Core Features & Use Cases - Adversarial Falsification: Treats every correctness claim as a hypothesis to disprove, attacking edge cases, malformed input, scale, concurrency, and security abuse paths. - Reproducible Evidence: Produces a clear verdict with reproducible failing cases and the exact verification commands that were run. - Regression Coverage: Adds a regression test for every defect found, complementing the engineer's test-first suite with boundary, negative, and cross-level tests. - Use Case: After an engineer ships a new API endpoint, use this Skill to attack it with null payloads, oversized inputs, and concurrent requests, then hand off a verdict with failing cases the engineer can reproduce. ## Quick Start Use the testing skill to adversarially probe the latest implementation for edge cases, malformed input, and security weaknesses, then report a verdict with reproducible failing cases.
