Claude Skills by yogsoth-ai

Multi-dimensional patent quality assessment — forward citations, family size, claim count, geographic breadth

Tactic: 迭代精化研究问题直到通过 FINER 5 项标准

Reframe research questions using abstraction laddering, HMW formulation, and Socratic probing. Find the most productive level and framing for investigation.

SOP: 综合所有中间产物产出最终研究问题集

Synthesize multiple dimension scores into radar chart data and compute overall readiness.

Random word/concept as thinking entry point (de Bono Random Entry)

Select random paper facet as creative stimulus. Uses genuine randomness in paper selection to break domain fixation.

Random word/paper/concept as thinking entry point. Use genuine randomness to escape fixation and open unexpected solution paths.

Use random word/concept injection as creative stimulus. Selects random concepts and forces connection to the problem space, generating unexpected solution paths.

Compare multiple ranking results to assess agreement and identify divergent items.

Produce the final ranking artifact from converged ratings and consistency report.

Strategy: 快速粗筛——两轮过滤将大量 gaps 压缩为可精排的候选集

Incorporate a new judgment into the rating model and return updated ratings for all candidates.

Re-derive conclusions under a negated assumption, tracking where the derivation diverges from the original.

Re-evaluate S/O/D scores after mitigation measures are in place. Validates that mitigations actually reduce risk as expected.

Reassemble decomposed fragments into novel structures through systematic recombination of components.

Reassemble decomposed system fragments into novel structural arrangements that create emergent value.

Campaign: Systematic adversarial attack from military/intelligence/AI-safety traditions. Core question: Can systematic adversarial attacks find fatal flaws? Methods: UFMCS Red Team Handbook v9.0, CIA SAT, Anthropic Red Teaming, NIST AI RMF, Inie et al. 12-strategy taxonomy.

Compile all problem reformulation analyses into a coherent report with a recommended new problem definition.

Reframe the problem from 4 professional perspectives to reveal what each discipline would focus on.

SOP: 指定变量间关系的方向与形式

Assess how removable a constraint is with effort estimate and dependency analysis.

Design concrete removal steps for a constraint with timeline and resource needs.

Assess paper completeness against ML Reproducibility Checklist

Ensure experiment reproducibility through systematic environment and seed control

Verify result reproducibility via re-runs with different seeds and ICC comparison

Complete strategy book for the research engine. Contains all campaign descriptions, routing tables, pre-conditions, and orchestration rules. CC reads this before generating any Research Spec.

Campaign: 将假设细化为精确的、框架化的研究问题

Are resources sufficient? — Quantify compute, data, time, human, and financial resource constraints

Estimate resources, budget, and timeline using parametric, analogous, and three-point (PERT) estimation methods.

Quantify resource demand vs supply vs gap for each resource category

Statistically analyze collected results, verify reproducibility, and synthesize findings

Collect experiment outputs — metrics, logs, artifacts — into structured result set

Validate results through statistical testing, ROPE judgment, reproducibility re-runs, and final synthesis

Argue for rejected candidates using Devil's Advocacy, Dialectical Inquiry, and Adversarial Collaboration to ensure elimination was justified.

Systematically reverse positive statements to generate creative inversions. Produces reversed statements with initial associations.

How to make it worse? → reverse for solutions. Generate anti-solutions then invert to discover novel approaches.

Hostile reviewer perspective — find fatal flaws, logical gaps, and missing evidence in a solution.

Balance portfolio risk and return using Markowitz mean-variance, CVaR, Risk parity, and Kelly criterion methods.

Assess methodological bias using RoB2, PROBAST, or QUADAS-2 validated tools

Strategy: Action Priority matrix — classifies failure modes into H/M/L priority using severity-weighted scoring per AIAG-VDA 2019 Action Priority tables.

Design experiments to identify failure boundaries and robustness limits

Compute robustness index across scenarios with sensitivity analysis

Test conclusion robustness via multi-model convergence — enumerate assumptions, generate alternatives, compare results, flag fragile conclusions.

Select portfolios that perform well across multiple future scenarios using Minimax regret, Robust optimization, Scenario planning, and Info-gap methods.

Role-play as reviewer/practitioner/theorist/novice/competitor to generate diverse perspectives on a solution.

Drill from surface symptoms to root causes via 5 Whys, Ishikawa decomposition, and Current Reality Trees. Validates each causal link with literature evidence.

Decide whether to continue iterating or stop based on consensus score, round number, and stability.

Find and challenge domain's unquestioned beliefs. Systematic identification and productive violation of dogma.

Find domain's unquestioned beliefs. Systematic identification of dogma that constrains innovation.