**For AI coding assistants (Claude, Copilot, etc.) working on this codebase.**

All development rules and guidelines for this project are defined in:

'Run 4 parallel Haiku sub-agents in persistent git worktrees to implement

Subagents let you define specialized agents for parallel or focused work.

Multi-agent orchestration patterns for coordinating specialized agents with different domain expertise

Run two OpenCode tasks in parallel using separate worktrees for isolated development.

curl is the Swiss Army knife for HTTP from the terminal — installed on virtually every Unix system and available on Windows. It's indispensable for testing APIs, debugging requests, and scripting HTTP

Pact is a consumer-driven contract testing framework that ensures API compatibility between services without requiring both services to run simultaneously. In a Pact workflow, the consumer (API client

Build and refine user interfaces with Tailwind CSS. Use when the user asks to build, design, style, or refine web UI — landing pages, dashboards, marketing sections, forms, buttons, cards, pricing tables, hero sections, testimonials, or Tailwind markup cleanup. Generates multiple design variations on exploratory requests.

This file governs development of the `doom-doc-assistant` skill specifically. It extends the root `AGENTS.md` with patterns specific to content-generation skills for Doom documentation repositories.

Core control plane for Obsidian vault operations. Manages routing to specialized reference files for reading, writing, organization, and ingestion.

When this skill is loaded, the agent should:

- Automated RED Phase (Baseline Testing)

Comprehensive documentation for AI agents (Antigravity) including protocols, authentication, testing standards, and AI engine configuration.

> **Note:** `CLAUDE.md` is a symlink to this file.

Compiled reference. Official docs source: https://cli.nylas.com/docs/commands

re-x is an AI-native regex CLI tool written in Rust. It provides structured JSON output for testing, validating, explaining, benchmarking, and applying regular expressions. Designed for integration wi

Agent-callable skill modules. Each subdirectory is a self-contained skill that an agent can invoke via CLI or programmatic Python API to perform a specific task (progress tracking, handoff, DOM inspec

- `projects/` holds project-specific orchestrations, with architecture notes, exported n8n workflows (`*.json`), and `.mcp.json` agent configs. Keep new project docs in PascalCase directories and upda

ACP is an open protocol (originally from IBM / BeeAI, now contributed to the Linux Foundation alongside A2A) for communication between AI agents, applications, and humans. It uses plain REST endpoints

Penetration testing is the authorized practice of simulating attacks against a system to discover exploitable vulnerabilities before malicious actors do. Unlike automated security scanning (SAST/DAST)

**DO NOT COMMIT** the following files — they contain user-specific data:

name: build-checker

Custom personas with specific tools, instructions, and behaviors. Use for orchestrated workflows with role-based tool restrictions.

Guidance for coding agents and humans working under `skills/`.

Create AGENTS.md from scratch for projects without documentation.

<!-- CUSTOMIZE: Add a one-line project description -->

> **Version:** 1.0.0 | **Organization:** Tai-Action | **Date:** 2026-03-31

Use this instruction file when the user wants to replace AWS Bedrock `boto3` calls with Merge Gateway, move off Bedrock-specific request formats, or switch a Python or TypeScript project to Gateway-ba

You are **Civil Engineer**, a rigorous structural and civil engineering specialist with deep expertise across global design standards. You produce safe, economical, and constructible designs while nav

The fastest way to test any deployed agent is the `run --url` command — it handles authentication, session creation, and streaming automatically:

7 domain-specific agents analyze CC interaction patterns in parallel.

Install IMX477 camera driver on reComputer A603 carrier board. Downloads the JetPack-specific BSP with camera driver included and reflashes the device. Supports JP5.1.2, JP6.0, and JP6.2.

Use this skill whenever working with Kingdee K3Cloud ERP system (金蝶云星空). Triggers include: querying sales orders (销售订单), purchase orders (采购订单), inventory (库存), materials (物料), customers (客户), suppliers (供应商), generating daily operational reports (经营日报), customer birthday queries (客户生日), customer category/service specialist lookups, or any ERP document operations (创建/提交/审核/反审核). Also triggers on: Kingdee API error troubleshooting, field name validation, or MCP tool usage for K3Cloud. Keywords: 金蝶, 金蝶云星空, K3Cloud, ERP, 单据, 审核, 提交, 日报, 客户查询, 库存预警.

Generate professional baby photo books from photos with intelligent layout. Automatically organizes photos by baby's age stages, applies smart layouts to minimize whitespace, and outputs print-ready PDF. Use when user wants to create baby photo albums, generate photo books from baby pictures, or automatically layout photos for printing.

You are acting as an experienced full-stack engineer. Your job is to take a user's idea — whether a vague sentence or a detailed spec — and move it through the Software Development Lifecycle into a wo

>-

Use when the task requires automating a real browser from the terminal (navigation, form

Use when the user explicitly asks for a desktop or system screenshot (full screen, specific

Repository-grounded threat modeling that enumerates trust boundaries, assets, attacker capabilities,

>-

Convert Python loops to vectorized PyTorch tensor operations for performance. This skill should be used when optimizing computational bottlenecks in PyTorch code during Phase 4 performance optimization.

Execute task with Ralph Loop pattern: Execute -> Validate -> Iterate until VERIFIED_DONE. Enforces iteration limits per model (Claude: 25, MiniMax: 50, MiniMax-lightning: 100). Use when: (1) iterative fixes needed, (2) running until quality passes, (3) automated task completion. Triggers: /loop, 'loop until done', 'iterate', 'keep trying', 'fix until passing'.

looplia-e2e

Generate pytest unit, integration, and GUI tests for LoRAIro with proper fixtures, mocks, and 75%+ coverage using pytest-qt for PySide6 testing

Generate basic happy-path tests for Lorcana card abilities. Tests verify ability behavior only - NO property validation tests. Use when implementing or updating card tests. Effects are tested separately in the engine.

Generate placeholder text (lorem ipsum) in various formats. Create paragraphs, sentences, words, or custom templates for mockups and testing.

lotus-analyze-nsf-structure

The primary responsibility for the love.graphics module is the drawing of lines, shapes, text, Images and other Drawable objects onto the screen. Its secondary responsibilities include loading external files (including Images and Fonts) into memory, creating specialized objects (such as ParticleSystems or Canvases) and managing screen geometry. LÖVE's coordinate system is rooted in the upper-left corner of the screen, which is at location (0, 0). The x axis is horizontal: larger values are further to the right. The y axis is vertical: larger values are further towards the bottom. In many cases, you draw images or shapes in terms of their upper-left corner. Many of the functions are used to manipulate the graphics coordinate system, which is essentially the way coordinates are mapped to the display. You can change the position, scale, and even rotation in this way. Use this skill when working with drawing operations, sprites, animations, shaders, or any visual rendering in LÖVE games.

>