What problem does it solve? When translating equations from a textbook into Python code there is no reference implementation to diff against, so correctness must be proven by making the mathematics test itself. This Skill defines how to build that proof for every chapter of a fluid mechanics learning project. ## Core Features & Use Cases - V1–V7 evidence ladder: analytic exact solutions, sympy symbolic re-derivation and dimensional checks, convergence/order-of-accuracy studies, conservation-law residuals, cited published benchmarks, book-printed values, and limit/symmetry/invariance checks. - Tiered evidence requirements: CORE items need at least two independent evidence levels, NOTE items at least one, and every notebook-facing function at least a smoke test with a physical sanity assertion. - Ready-made assets: default tolerance table, a pytest test-file skeleton, a benchmark catalogue (Blasius, lid-driven cavity, Sod shock tube, Rayleigh–Bénard), a verification report format, and a structured failure loop with the most common physics-coding bug hypotheses. - Use Case: After implementing the Blasius boundary-layer solver for Chapter 9, use this Skill to write tests that check the wall shear against Howarth's 0.332057 benchmark, verify the ODE residual symbolically, and confirm the far-field limit — then produce the chapter verification report. ## Quick Start Ask the assistant to verify the chapter implementation using the verify-implementation evidence ladder and produce the verification report.
