发现 Skills

Use when you have a written implementation plan to execute in a separate session with review checkpoints

You carry pointers. When you set them down, they become real.

Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.

Analyze educational lessons using an emergent lens framework (AI Lesson Lens Assessment). Evaluate lesson quality through 7 interacting lenses, synthesize an Emergence Footprint, and assign S-D tier grades. Use when asked to analyze lessons, evaluate teaching quality, review educational sessions, or assess courses. Supports transcript-only and transcript+video analysis modes.

Master competitive analysis templates including competitor deep-dives, battle cards, feature comparison matrices, and win/loss analysis. Use when analyzing competitors, creating battle cards for sales, positioning against alternatives, tracking competitive moves, or preparing for competitive threats. Covers competitive intelligence frameworks, positioning strategies, and templates from Crayon, Klue, and April Dunford.

Create a specification for non-trivial changes. Creates GitHub issue + spec file in a dedicated directory.

Build a structured documentary search strategy from a natural-language research question. Decomposes concepts, expands terminology (synonyms, broader/narrower terms, related terms), and produces 8–15 validated bilingual (EN/FR) search queries as strict JSON. Use this skill at the very start of any literature review or retrieval task, before calling any retrieval skill. Trigger on phrases like "build a search strategy for", "find search terms for", "systematic review on", "what should I search for", "generate queries about", or any request that implies going from a research question to searchable expressions. Always use this skill before search-works-openalex, search-records-sudoc, or any other retrieval skill.

Updates core project docs from session work. Activates on '/document', 'document this', 'update the docs', 'let's document', 'mark this off'. Reads session manifest, updates CHANGELOG/SPEC/STATE/BugTracker, enforces naming conventions, marks completed items done. Mid-session safe — not a closeout ceremony.

Assist Claude in running PyWGCNA through omicverse—preprocessing expression matrices, constructing co-expression modules, visualizing eigengenes, and extracting hub genes.

Guide Claude through SCSA, MetaTiME, CellVote, CellMatch, GPTAnno, and weighted KNN transfer workflows for annotating single-cell modalities.

Multi-agent debate system. USE WHEN council, debate, perspectives, agents discuss. SkillSearch('council') for docs.

Extract system improvements and monitor the Anthropic ecosystem. Use when planning upgrades, checking Anthropic, or tracking new Claude features.

Prompt injection testing. USE WHEN prompt injection, jailbreak, LLM security, AI security assessment, pentest AI application, test chatbot vulnerabilities.

Use OpenAI Codex CLI for coding tasks. Triggers: codex, code review, fix CI, refactor code, implement feature, coding agent, gpt-5-codex. Enables Clawdbot to delegate coding work to Codex CLI as a subagent or direct tool.

Query National Archives resources for historical governance records, institutional documentation, and long-term operational context.

Query public Estonian court-proceedings and related judicial information from Riigi Teataja court information pages.

Validate idea/project structure, documentation completeness, and readiness for next phase

Format Markdown drafts into styled HTML for preview, or publish directly to WeChat Official Account (GZH) draft box. Supports multiple visual themes with smart topic-based theme selection.

Specialist in designing and implementing scalable modular monolith architectures using NestJS with DDD, Clean Architecture, and CQRS patterns. Use when building modular monolith backends, designing bounded contexts, creating domain modules, implementing event-driven module communication, or when user mentions "modular monolith", "bounded contexts", "module boundaries", "DDD", "CQRS", "clean architecture NestJS", or "monolith to microservices". Do NOT use for simple CRUD APIs, frontend work, or general NestJS questions without architectural context.

Use when user says "find calls about [topic]", "get transcript for [meeting]", or "show calls with [person]". Automatically searches and analyzes Fireflies.ai call recordings. Extracts transcripts, summaries, action items, and meeting insights.

Deploy and manage applications, databases, and services on Coolify. Use when a user asks to deploy an app to Coolify, manage Coolify deployments, check deployment logs, sync environment variables, create or manage databases on Coolify, trigger a deployment, manage Coolify servers, or troubleshoot Coolify issues. Covers the Coolify CLI and REST API. For infrastructure provisioning on Hetzner, see hetzner-cloud.

When the user wants to create or optimize popups, modals, overlays, slide-ins, or banners for conversion purposes. Also use when the user mentions "exit intent," "popup conversions," "modal optimization," "lead capture popup," "email popup," "announcement banner," or "overlay." For forms outside of popups, see form-cro. For general page conversion optimization, see page-cro.

Assists with building, training, and deploying neural networks using PyTorch. Use when designing architectures for computer vision, NLP, or tabular data, optimizing training with mixed precision and distributed strategies, or exporting models for production inference. Trigger words: pytorch, torch, neural network, deep learning, training loop, cuda.

Implement approved plans with precision and quality for Linus Dashboard Home Assistant integration. Use when implementing features, executing approved plans, writing code following standards, or developing new functionality.