481 skills · sorted by relevance
Track machine learning experiments with Trackio by recording training measurements, viewing live dashboards, and flagging diagnostic events. Retrieve metrics and alerts through command-line queries, with JSON output for automated monitoring and experiment review.
Automate Google Workspace tasks through a command-line tool: read email, calendars, files, documents, and spreadsheet data, or make requested updates. Inspect authentication, return structured JSON data, and restrict allowed operations for agent workflows.
Find input files in confirmed workspace folders, save outputs to explicit locations, and check the results against task requirements. Useful in restricted execution environments where familiar folders may be missing or inaccessible.
Turn raw security findings into a ranked list for engineering review. Check claims against source code, merge duplicates, assess how feasible exploitation is, and assign component owners, with results in JSON and Markdown.
Search Rhea, a biochemical reaction database, using reaction IDs, compound names, enzyme classification numbers, or reaction terms. Get concise summaries of focused search results, with raw JSON available when you need machine-readable output.
Retrieve concise Genebass summaries linking a selected gene's variant group to traits across a broad association scan. Specify one Ensembl gene ID, choose a variant group, and control result counts, with optional JSON output.
Retrieve public eQTL Catalogue association data, linking genes and genetic variants, along with study metadata. Filter requests to specific identifiers and receive concise Markdown summaries or raw JSON when machine-readable output is requested.
Search Google images through Serper.dev and collect candidate image URLs, page links, titles, and source domains. Choose URL-only output or save raw JSON results, with country and language preferences available.
Turn JSON data into browser-ready infographics with metric cards, grouped bar charts, process diagrams, or mixed dashboards. Outputs include all styling and SVG graphics in one HTML file, with preset palettes and icons and no external dependencies.
Analyze Excel and CSV files to summarize statistics, filter records, compare groups, and join data across sheets or files. Export results as CSV, JSON, or Markdown tables for further work or sharing.
Audit breadcrumb structured data, a machine-readable description of a page’s navigation path. Check that BreadcrumbList matches the visible trail and site hierarchy, then guide JSON-LD changes using the rendered HTML and HTTP response.
Extract text and its page coordinates locally from PDFs, Office documents, and images. Use OCR to read scans, export structured JSON or Markdown, process document folders, and render pages as PNG images for agents that analyze visual content.
Convert local images, audio, video, documents, and data using tools installed on the computer. Preserves original files, keeps conversion local, and checks outputs for content, format, and relevant quality changes before reporting the result.
Review unread Gmail messages in a summary showing sender, subject, and date. Adjust the message limit, apply Gmail search queries, and include labels, with table or JSON output and no changes to the mailbox.
Improve prompts through test cases and baseline comparisons, measure how well AI answers use retrieved sources, and check agent configurations. Includes Python tools for prompt analysis, retrieval quality reports, and token estimates, with costs calculated from supplied rates.
Create realistic sample data for testing, development, and demos. Specify fields, record counts, value ranges, and business rules, then receive CSV, JSON, SQL INSERT statements, or a Python generator script with validation and usage instructions.
Choose and operate KrillinAI CLI commands for subtitles, dubbing, video rendering, covers, and speech. This skill guides setup and helps interpret structured responses, output records, generated files, and errors.
Fine-tune language models with TRL using GRPO, a method that compares scored responses to the same prompt. Covers training data, scoring functions, GPU-conscious settings, and monitoring for tasks with measurable correctness or required output formats.
Move Python execution to saved script files when execute_code_sandbox repeatedly fails or becomes unstable. The skill covers writing single or multiple files, running them through run_shell, inspecting outputs and errors, and removing temporary scripts afterward.
Look up RNA entries in RNAcentral by identifier or targeted search, retrieve cross-references to related database records, and receive concise summaries. Raw JSON output is also available when requested for further processing.
Retrieve public collection and dataset metadata from CELLxGENE Discover and turn it into concise summaries. Use targeted collection lookups to review single-cell research data, with raw JSON available when machine-readable output is needed.
Check GenLayer intelligent contracts before testing. This skill uses the GenVM linter to find forbidden code patterns, SDK usage errors, and type mismatches, with JSON output for repeated fixes and commands to extract contract interface definitions.
Look up Claude Code commands and options for interactive work or automation. Covers non-interactive execution, session management, output formats, tool permissions, and agent definitions, with command examples for scripts and CI/CD workflows.
Pick an LM output format per (task x consumer x model) rather than by reflex: different formats carry different cognitive load (e.g. code-in-JSON makes the same model write worse code than plain-text+diff, while asking for prose when you need a typed object fails the other way). Use when designing or debugging an LM's output schema, choosing between plain text / diff / JSON / tool-call / grammar-constrained output, or when a model's quality drops after wrapping its output in a structured format.
The English summary is being prepared. Showing the original description.