What problem does it solve? When a PyPTO operator fails precision verification—producing wrong outputs, numerical deviations, or inconsistent results—developers need a structured way to check user-level code and try known workarounds before suspecting framework bugs. This Skill provides that systematic troubleshooting workflow. ## Core Features & Use Cases - Frontend and Syntax Checks: Verifies use of the recommended pypto.frontend.jit decorator and audits dtype, shape, and valid_shape configuration in user code. - Prioritized Workaround Table: Applies seven ranked mitigations including avoiding view+reshape inplace=True, unroll_list=[1], submit_before_loop=True, the +0.0 trick, shape adjustment, pypto.zeros initialization, and valid_shape parameters. - Escalation Discipline: Routes unresolved cases to the pypto-precision-compare skill for bisection, and only reports a possible framework issue after all steps are exhausted. - Use Case: Your PyPTO Add operator passes on most shapes but fails when the tail axis is 1. The Skill walks you through frontend checks, then targeted workarounds like shape adjustment and valid_shape fixes until precision is restored. ## Quick Start Ask the assistant to troubleshoot your PyPTO operator whose precision verification fails, describing the failing shapes and symptoms so it can apply the step-by-step debugging workflow.
