Browse reusable Agent Skills, each with a clear purpose and practical guidance.
Use when the identification strategy is the bottleneck for a 《世界经济》 (The Journal of World Economy) manuscript — external shocks (tariffs, trade agreements, exchange-rate reforms, FDI negative-list) as quasi-experiments, plus shift-share / Bartik and gravity designs. Stress-tests exogeneity and inference, including the modern shift-share critiques.
Use when designing or auditing the empirical section of a KDD paper, where evidence combines quality deltas with scalability and efficiency measurements, temporal-leakage-safe splits, mechanism-isolating ablations, tuning-symmetric baselines, and, for the ADS track, post-launch measurement design that survives the desk check.
Use when positioning a KDD submission against the data-mining lineage (prior KDD volumes, ICDM, SDM, WSDM, CIKM, WWW) and the ML flagships, handling venue misattribution traps, cross-cycle resubmission overlap, concurrent arXiv work, and the mechanism-contrast style of novelty argument that ACM SIGKDD reviewers expect.
Use when hardening the reproducibility story of a KDD paper, where reproducibility is an explicit decision factor for area chairs. Covers claim-to-evidence tiers for mining pipelines, seeds and splits on large graphs and streams, baseline-tuning disclosure, compute reporting, and honest limits for ADS deployment results.
Use when defending the empirical design of a Language (LSA) manuscript on the terms of its subfield — elicitation and fieldwork, corpus construction, phonetic measurement, experiment, or the diachronic/typological sample. Language judges each kind of evidence by its own standards, and the design must support the theoretical claim. Defends the design; it does not run the analysis.
Use when anticipating how a Language (LSA) manuscript will be judged — the double-anonymous review, the general-audience and cross-framework bar, the desk-return filters (descriptive data dump, single-framework parochialism, undocumented data), and the decision categories. Sets expectations and stress-tests before submission; it does not write the paper.
Use when targeting Laser & Photonics Reviews or deciding whether a laser-physics / photonics review or original-research manuscript fits this venue. Encodes the journal's fit, the synthesis-and-significance bar, review and original-research rigor, house style, the review-vs-primary-research routing, official-submission re-check, and desk-reject heuristics.
Use when targeting Linguistic Inquiry (MIT Press) or deciding whether a generative-linguistics manuscript fits this venue. Encodes the journal's fit in generative syntax and its interfaces, the formal-analysis bar, squib format, house style, glossing and double-blind norms, official-submission re-check, and desk-reject heuristics.
Use when targeting Living Reviews in Relativity (LRR) or deciding whether an invited, continuously updated authoritative review in relativity and gravitation fits this very-high-impact Springer venue that publishes no primary research. Encodes the journal's fit, framing, evidence bar, house style, official-submission re-check, and desk-reject heuristics.
Use when the mathematical core of a Mathematical Finance (Wiley) manuscript is the bottleneck — adapted for a theory journal, this means assumptions, theorem statements, proof architecture, and generality, not causal/empirical identification. Stress-tests rigor before exposition is polished.