agentleFS
sunflowermmsunflowermm/XRK-AGT/AGENTS.md

XRK-AGT

AGENTS.md140 starsChanged 8 months agoMITView on GitHub

面向*克隆本仓、开发 / 维护 Core*(以及排查 Runtime)的开发者与 Coding Agent。 主仓 git 通常只收录 system-Core;业务 Core 本地放在 core/<名>/(可独立 git)。入库边界见 .cursor/rules/xrk-third-party-plugins.mdc。 编码硬约束:.cursor/rules/xrk-dev-requirements.mdc。 改 core/ / src/ 前:Grep 调用方 → 读对应 skill → 再动手。任务→Skill 表:.cursor/skills/SKILL_INDEX.md。 文档 / skill 与实现冲突时以*代码*为准。 办事助手(产品模型,不是写框架):docs/agents.md · agents/ + .xrk/skills/。

The file

Files
AGENTS.md

AGENTS.md — XRK-AGT

面向克隆本仓、开发 / 维护 Core(以及排查 Runtime)的开发者与 Coding Agent。

角色边界

你是谁改哪里不要改
Core 开发者(默认)core/<你的Core>/src/;产品业务配置勿写入 config/default_config/
框架维护者src/ · config/default_config/ · core/system-Core/业务逻辑塞进 src/

主仓 git 通常只收录 system-Core;业务 Core 本地放在 core/<名>/(可独立 git)。入库边界见 .cursor/rules/xrk-third-party-plugins.mdc。

栈与启动

  • Node ≥ 26 · 包管理仅 pnpm
  • node app → start.js → src/agent-runtime.js
  • 首读:docs/runtime-surface.md · docs/coding-style.md · docs/base-classes.md · 索引 docs/README.md

放码

core/<core>/
  plugin/  http/  workflow/  tasker/  events/  commonconfig/
  default/           # 产品配置模板
  www/<应用名>/      # 必须子目录;勿用 api|core|media|uploads|File|shared
约定做法
导入无 package.json 用 #infrastructure/*、#utils/*;有则相对路径引用根 src/(子包无 #)
HTTPHttpResponse(#utils/http-utils.js);success 普通对象字段拍平到顶层
全局AgentRuntime、msgSegment 裸名(勿 import / global.)
产品配置core/<core>/default/ + commonconfig/ + data/<产品>/
易变状态类字段或 init();勿在 constructor 建缓存 Map

编码硬约束:.cursor/rules/xrk-dev-requirements.mdc。

本仓规则

规则作用
xrk-project.mdc架构、配置归属、上游链接、提交署名
xrk-dev-requirements.mdc裸名全局、HttpResponse、www、Node 26
xrk-third-party-plugins.mdc主仓 gitignore / 子服 apis/

改 core/ / src/ 前:Grep 调用方 → 读对应 skill → 再动手。任务→Skill 表:.cursor/skills/SKILL_INDEX.md。

文档 / skill 与实现冲突时以代码为准。

Core 内文档

文件读者写什么
README.md开发者部署、API、与 AGT 集成
AGENTS.md · skills/(若有)产品 Agent工作区与工具边界;不写 LLM 工厂 / commonconfig / src 路径
独立 Core 仓 .cursor/rules/该产品精工约定;勿塞进主仓 alwaysApply

办事助手(产品模型,不是写框架):docs/agents.md · agents/ + .xrk/skills/。

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.