s-work

Falkicon 的头像作者 Falkicon

适用于 Mechanic/Fen 生态系统工作的通用技能。涵盖生态系统背景、开发工具、组件概述和 AFD 原则。用于日常插件工作、探索或不确定使用哪种专业技能时。触发器:工作、构建、开发、创建、机械、fen、fencore、fenui、生态系统、插件。

📖 中文导读:适用于 Mechanic/Fen 生态系统工作的通用技能。涵盖生态系统背景、开发工具、组件概述和 AFD 原则。用于日常插件工作、探索或不确定使用哪种专业技能时。触发器:工作、构建、开发、创建、机械、fen、fencore、fenui、生态系统、插件。 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。

name: s-work description: > General-purpose skill for Mechanic/Fen ecosystem work. Covers ecosystem context, development tools, component overview, and AFD principles. Use for everyday addon work, exploration, or when unsure which specialized skill to use. Triggers: work, build, develop, create, mechanic, fen, fencore, fenui, ecosystem, addon.

Fen Ecosystem

Expert guidance for working in the Mechanic/Fen WoW addon development ecosystem.

The Reload Loop (MANDATORY)

After ANY addon code change, you MUST verify the changes in-game:

  1. Ask the user to /reload in WoW (or trigger via keybinding CTRL+SHIFT+R)
  2. Wait for the user to confirm the reload is complete
  3. Then use the addon.output MCP tool (agent_mode=true) to get errors, tests, and console logs

CRITICAL: Do NOT call addon.output immediately after changes. The timing between reload and SavedVariables sync is unpredictable. Always wait for user confirmation before pulling output.

Ecosystem Components

Component Purpose Key Tools
Mechanic Development hub env.status, addon.output, reload.trigger
FenCore Pure logic library fencore.catalog, fencore.search, fencore.info
FenUI UI widget library Layout, Panel, Tabs, Grid, Buttons
MechanicLib Bridge library RegisterAddon, Print, RegisterTest

Essential MCP Tools

Task MCP Tool
Get Addon Output addon.output (agent_mode=true)
Lint Code addon.lint
Run Tests addon.test
Search APIs api.search
Search FenCore fencore.search
Env Status env.status

AFD Core Principles

Mechanic follows Agent-First Development (AFD) — github.com/Falkicon/afd

  1. Tool-First: All functionality must exist as an MCP tool before being added to any UI.
  2. Structured Results: All tools return predictable JSON schemas with success, data, and error.
  3. Agent Mode: Use agent_mode=true for distilled, AI-optimized output.

Routing Logic

Request type Load reference
Component details, architecture references/components.md
Daily workflows, patterns references/workflow.md
Quick API reference references/quick-api.md
MCP tool reference ../using-mechanic/references/afd-commands.md