What problem does it solve? Persistent-homology pipelines can run end-to-end and produce plausible-looking persistence diagrams that are quantitatively wrong, and internal consistency checks alone will not catch the error. This Skill anchors topological results to external published benchmarks before they are trusted in papers or used as downstream inputs. ## Core Features & Use Cases - Benchmark Anchoring: Compares results against published references such as Gidea-Katz 2017 financial pre-crash L1/L2 norm trends or the project's validate-topology benchmark table. - Diagram Sanity Checks: Verifies off-diagonal features are real rather than numerical noise, H0/H1/H2 counts are plausible, and total persistence scales sensibly with the filtration. - Metric Direction and Replication: Confirms W2/landscape distances move in the expected direction under controlled perturbations and that results reproduce across eras and cohorts. - Use Case: A researcher computes a persistence diagram for a financial time series and needs to confirm the W2 distance trend matches known pre-crash behavior before citing it in a paper. ## Quick Start Review this persistence diagram and Wasserstein distance result against published benchmarks and tell me whether it matches, deviates, or has no available benchmark.
