Skill UI

This skill captures and persists critical user decisions, project constraints, and architectural choices across multiple sessions and agents. It saves this context into a structured memory database (`mdbase`), ensuring that future development steps or agents can retrieve the original intent without re-deriving it from scratch. Ideal for documenting project rules, tech stacks, and design rationale.

A comprehensive guide for implementing view transitions in React and Next.js. Learn how to animate UI states, shared element animations, and route changes using the native ViewTransition API. Covers best practices for maintaining spatial continuity and adapting to different navigation contexts.

This skill provides a direct, fast shell invocation to capture the current git diff (staged, working-tree, or specific file list) as raw text. It is designed to replace slower LLM agent steps that merely execute 'git diff' commands, significantly speeding up CI/CD workflows and meta-steps that require repository state analysis.

A meta-skill designed to monitor long-running processes, such as model fine-tunes or CI builds, executed within a detached tmux session. It captures output, automatically uses a sub-agent to diagnose failures, proposes a fix, and records the complete debugging process to memory. Ideal for intermittent failure debugging and continuous integration pipelines.

Designed to capture a fully functional, standalone static HTML file from any modern web application (React, Next.js, Vue, etc.). It resolves all computed styles and inlines CSS and images (as Base64) into the output, ensuring the captured view can be shared or archived without external dependencies. Supports both automated snapshotting via Puppeteer and interactive capture.

Provides comprehensive, programmatic control over tmux sessions. Users can list, capture output from, send specific keystrokes, and programmatically navigate panes/windows in remote, long-running interactive CLI sessions. Essential for monitoring AI code generation tools (like Claude/Codex) or managing complex multi-terminal workflows via scripting.

Atlas Ledger is a critical companion designed to provide lasting 'memory' to project contracts. It automatically processes detected deviations or 'drift' (e.g., from audits or reviews) and distills these instances into formal, actionable contract clauses (WHEN/DON'T/INSTEAD). These clauses are written to a persistent project memory file, Atlas.md. This mechanism ensures that lessons learned from failure are systematically captured, allowing future contract cycles to incorporate these rules, thereby strengthening overall project governance and reliability.

This skill applies a rigorous, evidence-first discipline to coding tasks, inspired by Fable-like best practices. It guides agents through a structured workflow: inspecting the workspace, tracking goals and findings, grounding all conclusions in verifiable evidence, and executing thorough verification gates before declaring completion. Ideal for complex development, debugging, or adapting cross-model prompt guidance into robust, verifiable project rules.

This skill guides you through a comprehensive, multi-phase interview to capture every necessary detail of a new project idea. It is designed to run BEFORE any code is written, forcing detailed specification gathering across purpose, features, data models, and technical constraints. The output is a clean, self-contained prompt.md file, ensuring subsequent development is based on solid, unambiguous requirements and preventing costly mistakes.

A powerful command-line tool for recording the macOS screen and capturing system audio simultaneously. Unlike QuickTime or standard `screencapture`, this tool captures raw display and system audio directly via ScreenCaptureKit, eliminating the need for complex virtual audio drivers (like BlackHole) or elevated permissions. It is designed for scripted content creation, automated demos, and complex recording scenarios where system audio is required via the CLI.

This skill generates a highly customized code review checklist for any pull request. Instead of using a generic template, it adapts the review depth and specific checks based on the target language (e.g., Python, TypeScript), the nature of the change (e.g., bug fix, refactor), and the associated risk level (low, medium, high, critical). It ensures comprehensive coverage, guiding reviewers to focus on critical areas like security, correctness, and architectural integrity.

Use Mailtrap Email Sandbox to capture and inspect outbound emails without sending them to real recipients. This is crucial for development, staging, and CI/CD environments. It allows developers to test email content, headers, attachments, and spam scores using dedicated APIs, SMTP settings, or SDKs, ensuring mail integrity before deployment. Ideal for automation and rigorous quality checks.