发现 Skills

MANDATORY: Use instead of `git filter-branch` - safer git-filter-repo with recovery

Production-grade WebSocket connection management with connection limits, user-to-connection mapping, ping/pong health verification, and automatic stale connection cleanup.

Guide for initializing and consolidating Daem0n-MCP across project structures

Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

Guide for optimizing MSBuild build parallelism and multi-project scheduling. Only activate in MSBuild/.NET build context. Use when builds are not utilizing all CPU cores, when looking to speed up multi-project builds, or when evaluating graph build mode. Covers /maxcpucount, project dependency graphs, graph build (/graph), BuildInParallel, and reducing unnecessary project references.

Unarchive a Claude Desktop Cowork session by finding and modifying the session JSON file. Use when the user wants to restore an archived Cowork session.

Help a user rehearse an upcoming customer visit by playing the customer role with high fidelity based on a stored Contact Profile, then provide a structured reflection after the rehearsal. Use this when the user says things like "Help me simulate a conversation with X" / "I want to practice" / "I have a meeting with X, rehearse with me" / "I'm visiting X next week, let's run it through"; when the Contact Profile Agent suggests moving into rehearsal; when the user is preparing for a specific customer meeting and wants to feel how that customer might behave and how their own approach might land; and when the user mentions the visit scenario (first visit / proposal review / business negotiation / executive briefing / crisis handling) and wants to do a run-through first.

Exact detection thresholds for identifying malicious network patterns including port scans, DoS attacks, and beaconing behavior.

Parse and query the Human Metabolome Database (HMDB) local XML for metabolite information, chemical properties, biological context, disease associations, spectral data, and cross-database mapping. No public REST API — primary access via downloaded XML (~6 GB). For drug-focused queries use drugbank-database-access; for live compound lookups use pubchem-compound-search.

Telegram Bot API integration with managed authentication. Send messages, manage chats, handle updates, and interact with users through your Telegram bot. Use this skill when users want to send messages, create polls, manage bot commands, or interact with Telegram chats. For other third party apps, use the api-gateway skill (https://clawhub.ai/byungkyu/api-gateway).

Complete PowerShell expertise system across ALL platforms (Windows/Linux/macOS). PROACTIVELY activate for: (1) ANY PowerShell task (scripts/modules/cmdlets), (2) CI/CD automation (GitHub Actions/Azure DevOps/Bitbucket), (3) Cross-platform scripting, (4) Module discovery and management (PSGallery), (5) Azure/AWS/Microsoft 365 automation, (6) Script debugging and optimization, (7) Best practices and security. Provides: PowerShell 7+ features, popular module expertise (Az, Microsoft.Graph, PnP, AWS Tools), PSGallery integration, platform-specific guidance, CI/CD pipeline patterns, cmdlet syntax mastery, and production-ready scripting patterns. Ensures professional-grade, cross-platform PowerShell automation following industry standards.

Cost-tier dispatcher that routes low-complexity, non-security-sensitive tasks to external LLM providers (Gemini, Claude, local models) while enforcing user-selected cost preferences, billing thresholds, and daily caps. Acts as a Phase 1 Gateway Layer skill alongside tool-execution-gateway. All external model outputs are tagged with EXTERNAL trust score (0.3) before entering memory-guard. Integrates with ipi-defender, policy-engine, artifact-verifier, and post-gemini-validator. NEVER routes EXECUTE, VALIDATE, or ASSESS phase tasks to external providers. NEVER auto-switches from a zero-cost model to a paid model without explicit user consent.

Interactive QA session where user reports bugs or issues conversationally, and the agent files GitHub issues. Explores the codebase in the background for context and domain language. Use when user wants to report bugs, do QA, file issues conversationally, or mentions "QA session".

Triage issues through a state machine driven by triage roles. Use when user wants to create an issue, triage issues, review incoming bugs or feature requests, prepare issues for an AFK agent, or manage issue workflow.

Take a markdown file of raw material and shape it into an article through a conversational session — drafting candidate openings, growing the piece paragraph by paragraph, arguing about format (lists, tables, callouts, quotes) at each step. Use when the user has a pile of notes, fragments, or a rough draft and wants help turning it into something publishable.

Web testing with Playwright, Vitest, k6. E2E/unit/integration/load/security/visual/a11y testing. Use for test automation, flakiness, Core Web Vitals, mobile gestures, cross-browser.

Assist with api testing helper operations. Auto-activating skill for API Development. Triggers on: api testing helper, api testing helper Part of the API Development skill category. Use when working with APIs or building integrations. Trigger with phrases like "api testing helper", "api helper", "api".

Powerful coding-agent toolkit that converts an LLM into a fully featured agent that works directly on your codebase.

A股模拟交易系统 - 数据爬取、存储、策略回测、模拟交易 / A-Share Paper Trading System - Data Fetch, Storage, Backtest, Simulation

Autonomous financial research agent for stock analysis, financial statements, metrics, prices, SEC filings, and crypto data.

Convert AutoCAD DWG files (1983-2026) to Excel databases using DwgExporter CLI. Extract layers, blocks, attributes, and geometry data without Autodesk licenses.

Query headlines and articles from a self-hosted FreshRSS instance. Use when the user asks for RSS news, latest headlines, feed updates, or wants to browse articles from their FreshRSS reader. Supports filtering by category, time range, and count.

中文引导式学习助手。学习助手、知识学习、学习计划、学习路线、概念讲解、复习备考、考试复习、费曼学习法、循序渐进学习、学科辅导、自学、教程、课程学习、教材学习、知识点总结、记忆卡片、闪卡、Anki、学习方法、一对一辅导、自测试卷、选择题、简答题。Chinese guided learning assistant with step-by-step concept teaching, quizzes, tests, Feynman technique, and reviews. Use when: (1) learning any topic/skill/subject in Chinese, (2) creating a study plan or learning roadmap, (3) explaining concepts step by step, (4) preparing for exams or reviews, (5) using Feynman technique or analogies to learn, (6) generating flashcards or quiz questions, (7) creating self-assessment tests with multiple choice and short answer, (8) any Chinese-language tutoring or self-study scenario. 适用场景：学习新知识、制定学习计划、概念讲解、复习备考、做练习题、知识总结、用类比解释抽象概念、费曼学习法、自测试卷。

TikTok content strategy and creation system with local analytics tracking. Use when user mentions TikTok videos, content ideas, hooks, scripts, posting strategy, or analytics review. Generates video concepts, writes hooks and scripts, builds posting schedules, and tracks content performance. All strategies are advisory - actual results depend on execution and algorithm factors.