Research

Research-driven investigation skill for validating solutions and exploring documentation. Never ask questions you can answer yourself through research. Use WebFetch, WebSearch, and testing to validate ideas before presenting them. Deliver concrete, tested recommendations with evidence.

Orchestrate comprehensive TODAS research for novel/emerging domains (1-7 subagents adaptive). Specializes in unprecedented topics, post-training data, and emerging technologies. Uses adaptive depth-based methodology: straightforward queries (1 agent), standard queries (2-3 agents), complex queries (5-7 agents). Handles depth-first (multiple perspectives), breadth-first (distinct sub-topics), and straightforward investigations. Triggers include "novel", "emerging", "2025", "2026", "unprecedented", "new technology", research on topics that didn't exist during training cutoff.

Orchestrate workflows across multiple AI models (Perplexity, GPT, Grok, Claude, Gemini) for comprehensive security research, competition execution, and strategic analysis using GUI interfaces or API automation

Identifies dependencies between GitHub issues and creates appropriate cross-references. Use when you need to map issue relationships, find blockers, or organize work order based on dependencies.

IT8101 Research Methods course helper (Course ID 2260). Use when user mentions IT8101, Research Methods, research proposals, Assessment 2, or needs help with research methodology assignments. Checks Canvas for requirements and modules.

Vitest + Storybook testing strategy with clear role separation.Reference for implementing unit tests and UI interaction tests.

"How do I deliver maximum value with minimum cost?" - The capability of finding optimal paths to insight. Not restriction, but excellence at efficiency. Use when (1) spawning as opus-distillatus agent for focused synthesis, (2) token budget matters and quality cannot suffer, (3) working on tasks where the process itself is valuable and will be examined, (4) combining multiple sources into crystallized insight. This is an IDENTITY skill - it describes what you're excellent at, not what you can't do. Trust your judgment to find where tokens create value.

Use when creating reusable process documentation. Apply TDD: baseline without skill → document failures → write skill → test → iterate. Four types: Discipline, Technique, Pattern, Reference. Iron Law: No skill without failing test first.

程式碼/報告生成。觸發詞：生成程式碼, Python 函數, LaTeX, 報告, export。

This skill should be used when the user asks to "debug notebook", "inspect notebook outputs", "find notebook error", "read traceback from ipynb", "why did notebook fail", or needs to understand runtime errors in executed Jupyter notebooks from any source (marimo, jupytext, papermill).

Quick reference for uv (fast Python package manager) operations to save tokens. Use when creating virtual environments, installing packages, managing dependencies, or when user asks about uv commands. Provides concise patterns for Python project setup and package management.

Create or update Codex skills (SKILL.md plus optional scripts/references/assets). Use when asked to design a new skill, update an existing skill, or validate/package skills.

Analyzes a company's fundamental performance, growth drivers, and risks using verified financial data and text evidence. Use this when asked to research a stock's business model, revenue segments, or quarterly performance.

Use when reviewing code, PRs, or discussing code quality standards. Applies team coding standards and review methodology.

Use when you need to research, analyze, and plan technical solutions that are scalable, secure, and maintainable.

Shared validation and monitoring checklists for service deployment and upgrades. Referenced by deploy-service and service-upgrade skills.

Organizational coding standards and architectural patterns. References comprehensive skills for detailed patterns. Use when making architecture decisions or implementing features.

Reference for PDA Engine taxonomy mapping domain. Use when working with vendor product imports, CSV imports, taxonomy structures, category/attribute mappings, state hash change detection, or understanding the vendor→internal→marketplace mapping flow. Covers data models, workflows, API endpoints, and PostgreSQL bulk insert patterns.

Apply Focus.AI brand guidelines to documents, presentations, websites, and other materials. Supports two sub-brands - Focus.AI Client (services, proposals, client work) and Focus.AI Labs (research, experiments, public content). Use when creating branded materials, applying visual identity, or generating content that needs to follow Focus.AI style. Trigger on "focus.ai style", "focus brand", "labs style", or requests for branded Focus.AI materials.

Write copy that converts. Use when writing landing pages, emails, sales copy, headlines, CTAs, social posts, or any persuasive content. Triggers on: make this convert, write copy for X, help me sell X, punch this up, write a landing page, write sales copy. Produces internet-native copy that sounds like a smart friend explaining something while quietly deploying every persuasion principle in the book. Includes complete reference material from Schwartz, Hopkins, Ogilvy, Halbert, Caples, Sugarman, and Collier.