plugin-architect
Skill321 starsChanged 7 months agoMITView on GitHub
构建 AI Skills/Plugins 的标准方法论 (Based on Anthropic Standard)
In this repository32 other files it gives its agents.
- auto-page-syncSkill
- backlink-discoverySkill
- backlink-intelligenceSkill
- data-scraper-intentSkill
- favicon-icon-generatorSkill
- gemini-thinking-protocolSkill
- google-trends-to-pagesSkill
- html5-game-radarSkill
- itchio-new-game-huntSkill
- keyword-competition-analysisSkill
- multi-game-codes-hubSkill
- nextjs-seo-boosterSkill
- nextjs-seo-foundationsSkill
- null-axiom-twitterSkill
- python-agent-engineSkill
- roblox-game-data-scraperSkill
- roblox-hit-evaluatorSkill
- roblox-homepage-ranking-auditorSkill
- roblox-site-architectSkill
- rpg-stat-catalystSkill
- seo-auditorSkill
- seo-autopilotSkill
- seo-backlink-submitterSkill
- seo-link-strategySkill
- signallayer-backlinks-clientSkill
- site-keyword-researchSkill
- x-demand-radarSkill
- youtube-content-genSkill
- youtube-game-keywordsSkill
- youtube-game-scoutSkill
- youtube-intelSkill
- youtube-transcribeSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 39
Frontmatter: Must include name and description.
- Distinctive: line 84
Write strict "When This Skill Applies" triggers.
- Very specific: line 88
If user needs frequent access, create .agent/commands/<skill-shortcut>.md.
- Very specific: line 91
Ask AI: "Does this skill structure match the plugin-architect standard?"
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
🏗️ Plugin Architect Protocol
本 Skill 定义了本项目构建 AI Plugins 的唯一标准。所有 Skill 的创建与修改必须遵循此方法论。
When This Skill Applies
- 当用户要求"创建一个新 Skill"时。
- 当需要重构现有 Skills 以符合标准时。
- 当需要为 Skill 添加外部工具 (MCP) 或快捷指令 (Slash Commands) 时。
Instructions
一个完整的 Plugin (Skill System) 由四个核心组件构成。
1. The Identity (plugin.json)
插件的身份证。定义插件在系统中的唯一标识。
- Location:
.agent/.claude-plugin/plugin.json(Global) 或skill-name/plugin.json(Local) - Schema:
{ "schema_version": "1.0", "name": "devil-hunter-nexus-plugin", "description": "The cognitive engine and SEO factory for Roblox game sites.", "version": "1.0.0", "authors": ["Devil Hunter Team"] }
2. The Logic (SKILL.md)
AI 的思维链。必须包含明确的触发条件 (Trigger) 和结构化指令。
- Location:
.agent/skills/<skill-name>/SKILL.md - Frontmatter: Must include
nameanddescription. - Mandatory Sections:
## When This Skill Applies: 明确的触发条件 (Contextual Triggers)。- Good: "When the user asks to analyze SEO metadata."
- Bad: "Use this for SEO."
## Instructions: 具体的操作步骤、原则或代码规范。## Actions(Optional): 可用的工具或脚本引用。
3. The Interface (commands/)
用户的快捷入口。定义以 / 开头的显式调用指令。
- Location:
.agent/commands/<command-name>.md - Format:
--- description: Audit the current page for SEO compliance argument-hint: [url or path] active-skills: [nextjs-seo-foundations] --- # /audit-seo 1. Check presence of Metadata (Title, Description). 2. Validate JSON-LD Schema. 3. Report missing 'canonical' tags.
4. The Tooling (.mcp.json)
外部能力的延伸。通过 Model Context Protocol 连接真实世界。
- Location:
.agent/.mcp.json - Usage: 如果 Skill 需要执行复杂脚本 (e.g., Python Script, Database Query),需在此注册 MCP Server。
Workflow: How to Build a New Skill
- Define Intent: 是"思维模式" (Thinking) 还是"具体操作" (Action)?
- Thinking -> 重点编写
SKILL.md的 Instructions。 - Action -> 重点编写
commands/和scripts/。
- Thinking -> 重点编写
- Create Directory:
mkdir .agent/skills/<skill-name> - Draft SKILL.md:
- Copy metadata template.
- Write strict "When This Skill Applies" triggers.
- Write verifiable Instructions.
- Register Command (Optional):
- If user needs frequent access, create
.agent/commands/<skill-shortcut>.md.
- If user needs frequent access, create
- Validation:
- Ask AI: "Does this skill structure match the
plugin-architectstandard?"
- Ask AI: "Does this skill structure match the
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

