475 skills · sorted by relevance
Review an Anthropic Cookbook notebook for clear explanations, code quality, technical accuracy, and practical usefulness. It combines automated checks, including exposed credential detection, with a scored report and specific recommendations tied to sections and lines.
Draft and revise scientific manuscripts while linking claims to verified evidence. The skill organizes source records, checks consistency across text and results, tracks reporting requirements, and reviews references and disclosures, with human verification and final approval.
Find academic papers, verify authors and publication details, and create BibTeX entries, a structured format for references. Use it to organize bibliographies, remove duplicates, and check missing fields or mismatches with a manuscript.
Plan and review fluid simulations with FluidSim, from solver selection and resource estimates to output analysis and restart checks. Examine numerical accuracy through grid and time-step comparisons, conservation checks, and explicit limits on computing resources.
Simulate peer review through five perspectives covering journal fit, methods, subject expertise, interdisciplinary implications, and counterarguments. Receive separate review reports, an editorial decision, and a revision roadmap, or check whether revisions address earlier comments.
Create reusable agent skills or improve existing ones through test prompts and user feedback. Compare output quality, execution time, and token usage against a baseline, then refine instructions and descriptions so the skill is invoked in appropriate situations.
Create consistent records for evaluating AI agents, including task success, invalid tool calls, token use, and actual costs. Produce CSV tables, database definitions, and JSON validation schemas to support reproducible comparisons and checks for performance declines.
Turn recurring tasks into reusable agent skills, or improve skills you already have. Draft instructions, compare test outputs with and without a skill, review quality and resource use, and refine descriptions so skills trigger for appropriate requests.
Explains how PluginEval combines structural checks and experimental model assessments into quality scores. Helps interpret low scores, identify improvement areas, set marketplace thresholds, and explain what evaluation badges actually represent.
Design and refine prompts for AI applications using examples, reusable templates, and structured outputs such as JSON. Test variations for accuracy, consistency, response time, and token usage to diagnose unreliable answers and formatting failures.
Check important factual claims in courses or supplied content against available materials and authoritative sources. Review numbers, definitions, and contradictions, then receive a concise report with evidence and suggested corrections, or have generated course pages checked before delivery.
Read, extract text from, split, merge, fill, or create PDFs while preserving originals. Check page counts, extracted text, and file sizes, and flag accuracy limits when scanned pages require optical character recognition.
Check spreadsheet formulas for errors, broken references, inconsistent calculations, and unit mismatches. Review a selected range, one sheet, or an entire financial model, then receive findings ranked by severity with suggested fixes.
Create and review CSV geofeeds, files that publish location information for public IP addresses. Check formatting, country and region codes, and privacy concerns, then review suggested improvements in an HTML report.
Create and analyze AI model evaluation experiments in Arize using the ax CLI, a command-line tool. Collect real model outputs, export results, compare evaluation scores, and identify examples where a new model performs worse.
Add a model page to the SGLang Cookbook using shared templates. Organize verified launch settings and measured benchmarks, create configuration files and documentation, and update navigation, vendor cards, and page validation.
Compare exchange-traded funds by fees, tracking accuracy, liquidity, and market-price differences from asset value. Use the China-focused framework and Python examples to screen products, examine sector rotation, and study portfolio allocation.
Build real-time detection for ransomware encryption by combining changes in data randomness with file modification rates, renames, and process activity. Define behavioral scores and response thresholds, then check detection accuracy and monitoring overhead.
Forecast sales, demand, or sensor readings with Google's pretrained TimesFM model, without custom training. Check hardware and dataset memory needs, include external factors such as prices or holidays, and obtain forecasts with uncertainty ranges.
Reduce the memory needed to run large language models with GPTQ quantization, which stores model weights at lower precision. Covers loading compressed models, quantizing your own models, choosing execution backends, and connecting to memory-efficient fine-tuning.
Design and refine prompts for language model applications. Choose useful examples, create reusable templates and system instructions, and compare prompt variations using accuracy, consistency, response time, and token usage to guide improvements.
Build safety checks for LLM prompts and responses using Meta's LlamaGuard moderation model. Classify potentially harmful content, block flagged messages, serve checks through an API, and integrate them with NeMo Guardrails.
Reduce GPU memory use by loading language models in 8-bit or 4-bit formats with bitsandbytes and HuggingFace Transformers. Includes configuration examples, QLoRA fine-tuning of compressed models, and optimizers that reduce memory needed during training.
Fine-tune large language models with LoRA and QLoRA while limiting GPU memory use. Guides training settings and adapter management, using small sets of task-specific weights that can be saved, switched, or merged into a base model.