发现 Skills

Unified codebase quality review: merge readiness verdict + maintainability (Clean Code) + docs-vs-code consistency. Use for code review, quality check, refactor check, outdated docs check, or merge/production readiness.

Fast pre-DFT prescreening with GFN2-xTB via the xtb binary. Use to (1) pre-optimize geometries before DFT, (2) triage VAE/generative molecules that are likely to fail SCF/optimization, (3) run optional cheap frequency checks, and (4) produce standardized prescreen QC flags and artifacts for qchem-workflow batch pipelines.

An example user-invoked skill that demonstrates frontmatter options and the skills/<name>/SKILL.md layout

Tencent Docs skill providing full Tencent Docs operation capabilities. Use this skill when users need to operate Tencent Docs, including: (1) creating various online documents (document, Word, Excel, slides, mind map, flowchart); (2) managing knowledge base spaces (create space, query space list); (3) managing space nodes and folder/folder structures; (4) reading document content; (5) editing smart sheets (smartsheet); (6) editing documents.

A/B Testing Plan Generator

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

Build MCP servers for LLM-external service integration. Use for FastMCP (Python), MCP SDK (Node/TypeScript), tool design, API integration, resource providers.

Use when a task fails, a user corrects the assistant, a capability is missing, or a recurring better pattern should be logged for future improvement in Codex.

Standards for creating, testing, and optimizing Agent Skills. Use when creating, improving, catching regressions, measuring trigger rates, or writing eval cases for any skill.

Configure GitHub Copilot with custom instructions. Use when setting up .github/copilot-instructions.md, customizing Copilot behavior, or creating repository-specific AI guidance. Triggers on Copilot instructions, copilot-instructions.md, GitHub Copilot config.

Expert guidance for Microsoft 365 and Copilot development using the Teams Toolkit MCP server. Access manifest schemas, a knowledge base, code snippets for Teams AI/JS/BotBuilder SDKs, and troubleshooting. Use this when building Microsoft 365 agents, Teams apps, Copilot extensions, or when working with Teams development SDKs.

Use when all implementation tasks complete and tests pass — presents structured options for merge/PR/cleanup to complete the development work.

Query Polymarket: markets, prices, orderbooks, history.

Unsloth: 2–5× faster LoRA/QLoRA fine-tuning, less VRAM.

Internal development skill for working on the AI Game Builder plugin itself. NOT for game generation — this is for modifying the builder's skills, MCP server, Godot plugin, hooks, docs, and tooling. Load this when continuing builder development. Triggers on: "improve the builder", "add a new skill", "fix the MCP server", "update the plugin", or any request about developing the AI Game Builder itself.

Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing models, datasets, Spaces, buckets, repos, papers, jobs, and more on the Hugging Face Hub. Use when: handling authentication; managing local cache; managing Hugging Face Buckets; running or scheduling jobs on Hugging Face infrastructure; managing Hugging Face repos; discussions and pull requests; browsing models, datasets and Spaces; reading, searching, or browsing academic papers; managing collections; querying datasets; configuring Spaces; setting up webhooks; or deploying and managing HF Inference Endpoints. Make sure to use this skill whenever the user mentions 'hf', 'huggingface', 'Hugging Face', 'huggingface-cli', or 'hugging face cli', or wants to do anything related to the Hugging Face ecosystem and to AI and ML in general. Also use for cloud storage needs like training checkpoints, data pipelines, or agent traces. Use even if the user doesn't explicitly ask for a CLI command. Replaces the deprecated `huggingface-cli`.

REPL-friendly data visualization and literate programming for Clojure with the Kindly convention

Structured logging and telemetry for Clojure/ClojureScript with tracing and performance monitoring

Generate and maintain architecture guidance for a repository. Explores the repo via LLM, reads directory structure, manifests, configs, and imports, then generates `.archeia/` docs, `AGENTS.md`, and `CLAUDE.md`. Maximizes determinism through structured templates, evidence grounding, and self-validation.

Surfaces relevant instincts during work. Use when starting a task to check if any learned behaviors apply.

Package Castella applications for distribution using ux, PyInstaller, or Nuitka. Create executables, macOS app bundles, and cross-compile for other platforms.

Instructions for debugging agentstack-server during development

Complete workflow for converting Korean/English academic survey Markdown into publication-quality LaTeX PDFs. Handles Obsidian artifact cleanup (wikilinks, mermaid), XeLaTeX compilation with Korean font support (ucharclasses + Noto Sans CJK KR), BibTeX citations, TikZ figure generation, tcolorbox example boxes, and iterative format review. Use this skill whenever the user wants to build a PDF from an academic markdown file, compile a LaTeX survey paper, or fix formatting issues in an existing academic PDF. MANDATORY TRIGGERS: LaTeX survey, academic PDF, survey PDF, build paper PDF, xelatex, build_latex.py, Korean academic paper, survey compilation

Distributed computing for larger-than-RAM pandas/NumPy workflows. Use when you need to scale existing pandas/NumPy code beyond memory or across clusters. Best for parallel file processing, distributed ML, integration with existing pandas code. For out-of-core analytics on a single machine use vaex; for in-memory speed use polars.