Latest Agent Skills

Drafts Gmail replies on behalf of a client, matching their tone and sign-off, with optional templates and approval-based sending.

Checks library loans and searches the National Library Board of Singapore catalogue with filters for availability, material type, collection, and location.

Reads, searches, and sends Telegram messages via a CLI, including inbox summaries, chat history, contacts, and group member lookups.

Guides writing a failing test before production code, iterating through red-green-refactor cycles, and fixing bugs test-first.

Conduct multi-pass web and academic research, cross-check claims, and produce a cited report saved to file and memory.

Verifies a code change before it lands by scanning the staged diff, running tests, and delegating to an independent reviewer.

Manage GitHub repositories end-to-end: clone, create, fork, configure settings, protect branches, set secrets, create releases, and run workflows using gh or git+curl.

Diagnose bugs by establishing a tight feedback loop, tracing root cause, and verifying fixes with regression tests before making changes.

Analyze repositories for lines of code, language breakdown, file counts, and code-vs-comment ratios using pygount, with guidance on using language servers for symbol navigation.

Fetches the transcript of a YouTube video as text or JSON using the video's caption track.

Guides a coding agent through the full GitHub pull request lifecycle: branch, commit, open PR, monitor CI, fix failures, merge, and clean up, using either the gh CLI or git plus curl.

Reviews local git changes and GitHub pull requests, leaving inline comments and formal reviews.

Extracts text from PDFs and scanned documents, choosing the cheapest method that works, from direct file reads to full OCR.

Track multi-step work with todo, plan mode, kanban, and missions, each suited to different lifespans.

Guides users through configuring a LaRuche node by reading its live API state and providing click paths for settings, including LLM providers, messaging channels, secrets vault, MCP server, and memory.

Reports current weather conditions and forecasts for a city using live data from wttr.in or meteorological services, with the source named.

Stores, retrieves, and curates persistent facts in a structured memory map, distinguishing between user facts and reference text.

Judges agent outputs against LaRuche's quality standards, returning a structured scorecard with scores, verdict, and actionable corrections.

Translates plain-language monitoring requests into deterministic watcher rule trees for the LaRuche system.

Add plugin tools, connect MCP servers, or chain existing tools in one script within the LaRuche agent environment.

Runs three parallel reviewers to find code reuse, quality, and efficiency issues in a diff, then applies safe fixes and reports risky ones.

Create, search, triage, and label GitHub issues using the gh CLI or REST API, with guidance on avoiding duplicates and managing the backlog.

Answers factual questions by searching the web, fetching primary sources, and cross-checking claims before citing them.

Schedule prompts to run later, once or on a repeating cron expression, with tools to create, list, and delete scheduled tasks.

Write a skill, or forge a script tool, so a solved problem survives.

Authenticates to GitHub using a token, SSH key, or the gh CLI, with guidance for troubleshooting common auth failures.

Tracks RSS/Atom feeds and blogs using blogwatcher-cli, with feed discovery, HTML scraping fallback, OPML import, and read/unread article management.

Offload sub-tasks to other agents, specialist models, or background jobs, with guidance on choosing the right method and verifying results.

Inspect local git state, commit changes, and branch into isolated worktrees for speculative work.

Run GGUF models locally with llama.cpp, including finding the right file on the Hugging Face Hub, installing, quantizing, serving, and using Python bindings.

Rewrites text to remove AI writing patterns and add a natural, human voice, based on Wikipedia's guide to signs of AI writing.

Renders text or images as ASCII art for terminal-friendly output, including banners, frames, QR codes, and weather.

Read, search, create, and edit notes in an Obsidian vault using filesystem operations.

Controls Philips Hue lights, rooms, and scenes through the OpenHue CLI, including pairing, discovery, and presets.

Builds disposable HTML mockups in multiple design variants so users can compare and choose between them.

Runs throwaway experiments to validate feasibility before building, producing a verdict with evidence.

Finds academic papers on arXiv, checks citation counts via Semantic Scholar, and generates BibTeX entries.

Writes an actionable markdown implementation plan with bite-sized TDD tasks, exact paths, code, and commands, without executing it.

Geocode addresses, find nearby places, compute routes and distances, and get timezones using free OpenStreetMap data with no API keys.

Reference for the Paddle billing API covering products, prices, transactions, customers, and subscriptions with request and response examples.

Identifies common creative thinking anti-patterns and provides detection methods and alternatives.

Enforces GDPR-compliant deployment practices for Google Cloud Run, including Frankfurt region verification, PII-safe logging, data anonymization, retention policies, and CSP configuration.

Guides setting up Drizzle ORM with Vercel Postgres and Vercel KV for a manufacturing analysis app, covering schema definition, queries, migrations, caching, sessions, and rate limiting.

Guides AI agents on working with a deprecated, read-only repository of Claude Skills, covering skill structure, development workflows, and upload automation.

Converts documentation websites, GitHub repositories, and PDF files into AI-ready skills for multiple LLM platforms.

Reference for the Agent Script language used in Salesforce Agentforce, covering syntax, deployment methods, and troubleshooting.

Guides feature extraction and preprocessing for time series data using aeon transformers, covering collection and series transformers with code examples.

Teaches Rust traits, generics, and type system concepts with code examples and best practices.

Captures missing screenshots for MicroSims by running a local capture tool against each simulation directory.

Guides analysis of Neuropixels recordings from raw data to curated units, covering preprocessing, motion correction, spike sorting, quality metrics, and export.

Provides guidelines and best practices for aggregates and aggregate roots, referencing detailed documentation.

Converts legacy agent resources such as subagents, commands, and rules into the unified Agent Skills format, with classification, conversion, and verification steps.

Guides the agent through structured protocols for writing technical specifications, implementation plans, and summaries, with file naming and task directory conventions.

Coordinates multi-service releases, manages container registries, and sets up CI/CD pipelines across platforms.

Predict continuous values from temporal sequences using aeon's time series regressors, covering convolutional, deep learning, distance-based, feature-based, hybrid, interval-based, and shapelet-based approaches.

Automates the preparation of Easy difficulty diagram specifications from a CSV report and chapter files, generating reports, execution plans, and spec files for MicroSim generation.

Test your understanding of MkDocs, Material theme, configuration, navigation, version control, and GitHub deployment with these questions.

Defines quality standards and validation checklists for agent workflows, Python tools, and documentation, including automated checks and pre-commit gates.

Provides Python bindings to interact with the DNAnexus platform, enabling file uploads, job management, and API calls.

封装阶段3的Copy Spec为可批量执行的提示词包，生成JSONL请求并调用APIMart API出图。