发现 Skills

Skill to clear Open-WebUI embedding cache to fix 500 errors caused by architecture mismatch.

Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-checking, hallucination detection, PII filtering, toxicity detection. Uses Colang 2.0 DSL for programmable rails. Production-ready, runs on T4 GPU.

This skill should be used when the user asks "what should I work on", "what's next", "pick a task", or when needing to select between multiple available tasks. Provides guidance for prioritizing and selecting tasks.

Use when Playwright scripts fail, tests are flaky, selectors stop working, or timeouts occur - provides systematic debugging approach for browser automation issues

Professional software implementation skill for building features, components, and systems through multi-phase TDD development and incremental delivery.

Systematic framework for resurrecting and modernizing legacy codebases through archaeology, resurrection, and rejuvenation phases. Activate on "legacy code", "inherited codebase", "no documentation", "technical debt", "resurrect", "modernize". NOT for greenfield projects or well-documented active codebases.

Expert advisor for code quality, design patterns, and refactoring based on the Refactoring Guru knowledge base. Use this skill when the user shares code and asks what is structurally wrong; mentions code smells (Long Method, Shotgun Surgery, Feature Envy, etc.); asks which design pattern to use (Factory, Observer, Strategy, Decorator, Command, etc.); wants to refactor messy code; asks about composition vs. inheritance; needs to eliminate duplication, untangle dependencies, or simplify conditionals. Also trigger on casual phrasings like “this code is a mess,” “I have to change 10 files every time I add a feature,” “this method does too much,” “should I use inheritance or composition?,” “what pattern fits this problem?,” or “how do I structure this so it’s easier to maintain?”. Use for DESIGN and REFACTORING questions — how code is structured or which pattern to apply — not for debugging runtime bugs, writing tests, or explaining language features.

Build scroll-driven WebGL + DOM sites with React Three Fiber and three.js, with progressive enhancement, Next.js integration, and animation stacks.

Add Gmail integration to NanoClaw. Can be configured as a tool (agent reads/sends emails when triggered from WhatsApp) or as a full channel (emails can trigger the agent, schedule tasks, and receive replies). Guides through GCP OAuth setup and implements the integration.

Triage and resolve Debian Linux issues with apt, systemd, and AppArmor-aware guidance.

Expertise in MLIR (Multi-Level Intermediate Representation) and CIR (Clang IR) development for domain-specific compilation and high-level optimizations. Use this skill when building ML compilers, domain-specific languages, or working with multi-level compilation pipelines.

Create, organize, and distribute content across Workspace.

Blazor UI development specialist. Use for building Blazor Server or WebAssembly apps, component development, state management, or form handling.

IdeaVim git workflow conventions covering commits, branches, PRs, and CI. Use when creating commits, managing branches, creating pull requests, reviewing git history, or any git-related activity in the IdeaVim project.

Four Pillars (BaZi) destiny analysis. Collects birth information interactively (name, former names, solar/lunar birth date, birth hour, gender, birthplace), then generates the Four Pillars chart and provides a professional analysis based on classical texts (Qiongtong Baodian, San Ming Tong Hui, Di Tian Sui, Yuanhai Ziping, Ziping Zhenquan, etc.). Use this skill whenever the user asks about 八字, 四柱, 命理, 算命, Bazi, fortune telling, birth chart analysis, or wants to know their BaZi chart, luck cycles, major luck (Da Yun), or annual fortune. Triggers include common phrases such as: calculate bazi, read bazi, bazi reading, generate bazi chart, four pillars, life chart, fortune telling, check my bazi, I want a bazi reading, analyze bazi, cast a chart, bazi analysis, fortune telling, birth chart, cast a hexagram, check luck, fate analysis. Even if the user only mentions “fortune telling” or “BaZi” without explicitly asking to use the skill, this skill should be invoked.

Interact with Zotero reference management libraries using the pyzotero Python client. Retrieve, create, update, and delete items, collections, tags, and attachments via the Zotero Web API v3. Use this skill when working with Zotero libraries programmatically, managing bibliographic references, exporting citations, searching library contents, uploading PDF attachments, or building research automation workflows that integrate with Zotero.

Use this skill when writing long-form blog posts, tutorials, or educational articles that require structure, depth, and SEO considerations

Lanhu app development capability for building and managing Lanhu-related application features

Feishu (Lark) Spreadsheets: create and manipulate spreadsheets. Create spreadsheets and write headers and data, read and write cells, append rows, search for cell contents in a known spreadsheet, and export spreadsheet files. Use when a user needs to create spreadsheets, perform bulk read/write operations, search within a known sheet, or export/download a spreadsheet. If the user intends to search cloud storage for spreadsheet files by name or keyword, use lark-doc's docs +search to locate the resource first.

Guidance for implementing differential cryptanalysis attacks on FEAL and similar Feistel ciphers. This skill should be used when asked to break FEAL encryption, recover cipher keys through differential attacks, or implement cryptanalysis techniques on block ciphers with weak round functions. Covers proper differential characteristic construction, not ad-hoc statistical methods.

Systematically investigates causal relationships to identify true root causes rather than correlations or symptoms. Distinguishes genuine causation from spurious associations, tests competing explanations, and designs interventions addressing underlying drivers. Use when investigating why something happened, debugging systems, analyzing failures, evaluating policy impacts, or when user mentions root cause, causal chain, confounding, spurious correlation, or asks "why did this really happen?"

Optimize for behavioral SEO signals including CTR, dwell time, engagement metrics, and Bing-specific factors.

Expert guidance for Odoo ERP development including Python ORM, XML views, and module architecture