Skills Catalog

Read, create, and modify Microsoft Word documents. Extract text, add content, format paragraphs, and work with document structure.

Work with Excel spreadsheets. Read data, create charts, write formulas, format cells, and analyze tabular data.

Extract text and images from PDFs. Read multi-page documents, analyze layouts, and work with scanned documents.

Create and edit PowerPoint presentations. Add slides, insert images, format text, and build slide decks.

Generate creative algorithmic artwork using code. Create patterns, fractals, and generative designs.

Design graphics using HTML Canvas. Create illustrations, diagrams, and visual content programmatically.

Create animated GIFs for Slack reactions. Design custom emojis and animated stickers.

Build interactive web artifacts and prototypes. Create HTML/CSS/JS demos that run in the browser.

Build Model Context Protocol servers. Create custom tools and integrations for Claude.

Test web applications automatically. Write and run end-to-end tests for web interfaces.

Create and maintain brand style guides. Define colors, typography, and visual identity systems.

Write internal communications and announcements. Create memos, newsletters, and team updates.

Build your own Claude Code skills. Package reusable prompts and workflows for others.

Design and build frontend interfaces. Create responsive layouts, components, and UI elements.

Generate custom themes and color palettes. Create consistent design systems for applications.

Structured workflow for collaborative document creation across gathering context, refining content, and testing stages.

Enforces the Red-Green-Refactor cycle: write failing test, verify it fails, write minimal passing code, verify tests pass, refactor while keeping tests green.

Structured four-phase root cause analysis: Root Cause Investigation, Pattern Analysis, Hypothesis Testing, Implementation. Always find root cause before attempting fixes.

Mandatory verification gate with 5-step process: identify verification command, run it freshly, read full output, verify claim, then make the claim with evidence.

Transforms ideas into validated designs through three phases: Discovery (clarify purpose, constraints), Exploration (propose approaches with trade-offs), Design (detailed architecture).

Create comprehensive implementation plans with exact file paths, complete code examples, specific test commands with expected outputs, and git commit instructions.

Structured methodology for implementing pre-written plans through batched task execution (approximately 3 tasks per batch) with architect review cycles.

Enables concurrent investigation of multiple unrelated failures by creating focused agent tasks that work simultaneously on different problem domains.

Guides through structured code review process by dispatching the code-reviewer subagent. Establishes when reviews are mandatory versus optional.

Structured approach to evaluating code review feedback through technical verification. Five-step process: read, restate, check conditions, evaluate, respond.

Creates isolated git worktrees with smart directory selection and safety verification. Three phases: Directory Selection, Safety Verification, Setup & Verification.

Guides through completing development work: Verify Tests Pass, Determine Base Branch, Present Integration Options (merge, PR, keep, discard), Execute, Cleanup.

Orchestrates execution through Plan Extraction, Per-Task Cycle (implement, test, self-review, spec compliance, code quality), and final code review.

Adapts TDD to documentation creation. Establish baseline agent behavior (RED), create minimal documentation (GREEN), systematically close loopholes (REFACTOR).

Foundational framework mandating relevant skills be invoked before any action. Decision flowchart: receive message, determine skills, invoke, announce, follow, respond.

Build and test iOS apps using the simulator. Create Swift/SwiftUI projects and debug on simulated devices.

Penetration testing and web fuzzing using ffuf. Discover hidden directories, files, and vulnerabilities.

Browser automation with Playwright. Write end-to-end tests, scrape websites, and automate workflows.

Create interactive data visualizations with D3.js. Build charts, graphs, and custom visual representations.

Scientific computing libraries integration. Work with NumPy, SciPy, and other scientific Python packages.

Generate favicons, app icons, and web assets. Create all the icons needed for web and mobile apps.