The library

evm-token-decimals · DeFi parsing tasks where token amounts must be normalized by contract decimals.

exa-search · Use when working with skill named exa-search.

revenue_attribution.skill · When evaluating which marketing campaigns actually drive client acquisition and transaction revenue.

hugging-face-community-evals · Compare model performance against community standards without implementing custom eval logic.

hugging-face-dataset-viewer · Inspect and validate Hugging Face datasets before training without downloading entire dataset.

hugging-face-papers · When analyzing or summarizing AI research papers with automatic metadata extraction and linking to related Hub artifacts.

campaign-analytics · When measuring campaign effectiveness across channels and deciding budget allocation for paid marketing or performance optimization.

hugging-face-trackio · When running long training jobs and need real-time visibility into convergence, hardware utilization, and early stopping signals.

hugging-face-vision-trainer · When fine-tuning vision models (detection/classification/segmentation) for domain-specific image tasks with automatic Hub integration.

cdo-review · When data architecture, AI training data, or data monetization decisions need scrutiny on consent/dependencies

skill · When using skill

confusion-matrix-generator · Verification workflows where you need both supporting evidence and known conflicts for a claim, e.g. deal history, research synthesis, coaching notes.

networkx · Understanding network topology and critical path analysis.

customer-success-manager · Scoring customer health and churn risk deterministically using weighted multi-dimension models without external ML.

data-quality-auditor · Systematically profiling datasets to surface silent nulls, outliers, and distribution shifts before they corrupt downstream analysis.

ml-engineer · Training, fine-tuning, and evaluating PyTorch models with experiment tracking and reproducible configurations.

optimize · Profiling code and infrastructure to reduce latency, memory, and costs without trial-and-error tuning.

retrieval · Measures recall and precision before/after every change—catches regressions that impact search quality invisibly.

osint · Authorization framework prevents unauthorized investigations—checks all compliance gates before any collection starts.

parser · Semantic parsing extracts meaning from messy human text—outperforms regex on variants and typos.

dossier · Assembles multiple sources into a narrative with timeline—shows how events connected in ways raw data alone cannot.

experiment-designer · Product teams running defensible experiments with clear success criteria and statistical stopping rules.

postgres-patterns · Writing production PostgreSQL queries and schemas with Supabase best practices for performance and security.

postgresql · When you need to model a relational schema with correct constraints, performance tuples, and MVCC semantics.

prediction-market-oracle-research · When you want to use market-implied probabilities to inform strategic decisions while controlling for liquidity, resolution rules, and manipulation.

tavily · When you need production-grade tavily at scale.

mviz · Use for mviz operations

blogwatcher · Monitor blogs and RSS/Atom feeds for updates using the blogwatcher CLI.

research-knowledge-core · running any research, literature, IP, genomic, codebase, or docs-lookup task — the discipline every spoke

litreview · : 'litreview on [topic]', 'literature review on [topic]', 'I'm starting a literature review on X', 'I'm w

loop · Incremental optimization loops where one change per run is tested against a metric

risk-metrics-calculation · Quantifying and monitoring portfolio, credit, or operational risk when stakeholders need numerical risk bounds

goplaces · Managing travel destinations and itineraries with programmatic queries and updates.

scientific-db-pubmed-database · Executing scientific-db-pubmed-database skill with targeted automation.

scientific-pkg-gget · First-pass genomic lookups that need reproducible logs and standard database access before committing to local BLAST pipelines.

scientific-thinking-literature-review · Building citation-backed background sections where systematic protocol and deduplication rules prevent claim gaps.

scientific-thinking-scholar-evaluation · Thesis chapters and proposal review where rubric consistency and evidence traceability matter more than speed.

segment-cdp · Performing specialized work in its domain.

performance-profiler · Performing specialized work in its domain.

product-analytics · When you need retention curves and cohort comparison for product validation instead of single-point metrics.

update-pricing · One-time pricing updates where approval gate and change tracking matter

spark-optimization · Data engineers squeezing performance from Spark jobs without rewriting logic

2506-keyword_search_c279eebe · Indexed keyword search beats browsing untagged prompt files.

sympy · When research requires exact symbolic solutions rather than numeric approximation, or closed-form derivations.

senior-data-scientist · Reproducible ML research where feature selection, experiment tracking, and cross-validation are non-negotiable.

senior-ml-engineer · Taking a trained model to production when operational monitoring and automated retraining are concerns

setup · Setting up reproducible optimization experiments with clear metrics and success direction.

predictit · Monitoring US political prediction markets and retrieving real-time pricing without needing an API key.

videodb · Find exact moments in long video by spoken content or visual description, then export clips

firecrawl-cli · Extract structured data from dynamic websites in CI/CD pipelines or local scripts without browser overhead