Nexus: Instructions file for Claude Code

CLAUDE.md

Nexus CLAUDE.md is an instructions file for Claude Code from MT-gar/Nexus. It costs 1,792 tokens per session, scanned A, original, MIT.

A project instruction file for Claude Code in the MT-gar/Nexus repository. It documents the project, its technology stack, common commands, and dependency setup.

In plain words
What is it for?
Use it when Claude Code needs to understand Nexus, install its dependencies, or run its usual commands.
Why use it?
It gives the coding agent project-specific guidance instead of making it infer how the repository works.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

This is MT-gar/Nexus's own configuration. It tells Claude Code how to work on Nexus itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Nexus configures →

View source ↗ MT-gar/Nexus
Reuse

Borrowing it

Nothing to install: this file belongs to MT-gar/Nexus. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/MT-gar/Nexus/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/MT-gar/Nexus

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for Nexus CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/mt-gar/nexus/claude-md/github.svg)](https://agentmods.dev/instructions/mt-gar/nexus/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/mt-gar/nexus/claude-md"><img src="https://agentmods.dev/badge/instructions/mt-gar/nexus/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for Nexus CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/mt-gar/nexus/claude-md"><img src="https://agentmods.dev/badge/instructions/mt-gar/nexus/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 1,792 This file is loaded in full into every session.
When invoked 1,792 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.01792 $0.01792
Opus 5.5 $0.00717 $0.00717
Sonnet 5.5 $0.00358 $0.00358
Haiku 4.5 $0.00179 $0.00179

Measured 1mo ago against content hash 90735f5352ee, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

Nexus CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 140 lines

How it starts

The opening of the file, as written. The whole thing — 140 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

项目概述

多 Agent 会话协调系统,由两部分组成:

  • MCP Server — 中枢协调服务器,多个 Claude Code、Codex 会话通过它实现跨会话通信、状态共享、任务调度
  • Coordinator Skill — .claude/skills/ 下的 canonical skill 文件,.agents/、.codex/ 提供客户端适配

技术栈

  • 语言: TypeScript (strict mode)
  • 运行时: Node.js 18+
  • 数据库: SQLite (better-sqlite3) — 存储会话、任务、消息
  • 协议: MCP (Model Context Protocol) — 官方 @modelcontextprotocol/sdk
  • 构建: tsx (开发) + tsc (生产构建)
  • 测试: vitest

常用命令

# 安装依赖
npm install

# 开发模式运行 MCP server
npx tsx src/server.ts

# 构建
npm run build

# 运行测试
npx vitest

# 运行单个测试文件
npx vitest run tests/session.test.ts

# 类型检查
npx tsc --noEmit

# Lint
npx eslint src/

# 运行软件研发全流程协同 8 步闭环 demo(缺陷聚合→RCA→修复→发布→复盘)
npx tsx scripts/demo-devflow.ts

架构

src/
├── server.ts          # MCP server 入口,注册所有 tools/resources
├── db.ts              # SQLite 数据库初始化和迁移
├── session.ts         # 会话管理 (register, heartbeat, status)
├── message.ts         # 消息传递 (send, get, ack)
├── task.ts            # 任务管理 (create, claim, update, list)
├── defects.ts         # 缺陷聚合与去重
├── approvals.ts       # 审批域(高风险变更强制人工审批)
├── releases.ts        # 发布与回滚(审批门禁 + 失败自动回滚)
├── retros.ts          # 复盘与知识沉淀(轻量 RAG 检索)
├── json.ts            # JSON 字段解析工具(SQLite 字符串 ↔ 对象)
├── types.ts           # 共享类型定义

.claude/skills/
├── coordinator/       # 通用多会话协调 skill(产品经理/规划师/开发/watchdog)
└── devflow/           # 软件研发全流程协同 skill
    ├── SKILL.md       # 入口:角色选择(6 个角色)
    ├── workflows.md   # 8 步闭环工作流
    └── agents/        # 每个角色的 SKILL.md(含 9 要素)

.agents/skills/
├── coordinator/
│   └── SKILL.md       # Codex 项目级适配入口
└── devflow/
    └── SKILL.md       # devflow 适配入口

scripts/
└── demo-devflow.ts    # 8 步闭环冒烟脚本(运行轨迹证据)

.codex/config.toml # Codex MCP 配置

tests/ ├── session.test.ts ├── message.test.ts └── task.test.ts


### 核心模块职责

- **session.ts** — 会话注册/注销、心跳维护、状态查询。每个客户端会话启动时调用 `register_session` 获取唯一 ID。
- **message.ts** — 异步消息队列。支持点对点消息和广播,消息有 TTL 和 ACK 机制。
- **task.ts** — 任务看板。支持创建/认领/更新/完成,任务有状态机(todo → in_progress → done/blocked),支持依赖关系和分配。
- **defects.ts** — 缺陷域。多源聚合 + 归一化 Jaccard 相似度去重(阈值可调),重复上报合并到规范缺陷并累积来源证据。
- **approvals.ts** — 审批域。高风险变更(high/critical)强制 rollback_plan;pending 超时自动 expired;approve/reject 记录审批人/时间/意见。
- **releases.ts** — 发布域。创建发布自动发起 high 风险审批;进入 testing 必须审批通过(服务端门禁);版本唯一性强制;回滚记录 target_version/reason(幂等保护)。
- **retros.ts** — 复盘与知识域。复盘报告(根因/影响/改进措施)+ 知识条目(root-cause/fix-pattern/rule/lesson 标签),关键词+标签轻量 RAG 检索。

### MCP Tools 清单

| Tool | 说明 |
|------|------|
| `register_session` | 注册新会话,返回 session_id |
| `heartbeat` | 会话心跳,维持活跃状态 |
| `update_status` | 更新会话状态(idle/working/error) |
| `list_sessions` | 列出所有活跃会话 |
| `send_message` | 向指定会话发送消息 |
| `get_messages` | 获取待处理消息 |
| `ack_message` | 确认消息已处理 |
| `wait_for_work` | 统一等待消息/任务并自动心跳 |
| `wait_for_message` | 长轮询等待新消息 |
| `create_task` | 创建任务(含 spec/todo) |
| `claim_task` | 认领任务 |
| `update_task` | 更新任务状态/进度 |
| `list_tasks` | 按条件查询任务 |
| `watch_task` | 订阅任务变更通知 |
| `wait_for_task` | 长轮询等待指定任务变化 |
| `query_audit_logs` | watchdog/observer 查询审计日志 |
| `ingest_defect` | 录入缺陷(多源聚合),自动相似度去重合并重复上报 |
| `list_defects` | 按条件查询缺陷列表(状态/严重级别/来源) |
| `update_defect` | 更新缺陷(定级/分诊/关联 RCA/修复任务/关闭) |
| `request_approval` | 创建审批请求(高风险变更必须含回滚计划) |
| `list_approvals` | 查询审批请求(默认待审批) |
| `approve_request` / `reject_request` | 人工审批通过/拒绝(仅人工会话可调用) |
| `create_release` | 创建发布(自动发起人工审批,强制回滚计划) |
| `update_release` | 更新发布状态(testing/verifying/live/failed),含审批门禁 |
| `rollback_release` | 回滚发布(失败自动回滚,记录目标版本与原因) |
| `list_releases` | 查询发布历史 |
| `create_retro` / `list_retros` | 创建/查询复盘报告 |
| `publish_knowledge` | 沉淀知识条目到研发知识库 |
| `search_knowledge` / `list_knowledge` | 检索/查询知识库(轻量 RAG) |

## 设计原则

- **会话身份**: 每个客户端会话通过 `register_session` 获得唯一 ID,后续所有操作携带此 ID
- **消息是异步的**: 发送方不阻塞,接收方按需拉取。消息持久化在 SQLite 中
- **任务是共享真相源**: 所有会话看到同一个任务看板,状态变更原子化
- **心跳检测**: 超过 1 小时无心跳的会话标记为 stale,其持有的任务自动释放
- **客户端适配**: `.claude/skills/coordinator/` 是工作流真相源;Codex 入口只做适配,不复制业务规则

Read the full file on GitHub · 140 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 1mo ago Changed · +45 lines · +908 tokens per session 90735f5352ee
  2. 2mo ago First seen · 95 lines · 884 tokens per session scan A 9b1a0dc26f1a

Subscribe to this mod's changes

Nexus CLAUDE.md is an instructions file published in the GitHub repository MT-gar/Nexus (2 stars, last pushed 1mo ago), licensed MIT. It adds 1,792 tokens to every session, about $0.0072 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 9,715 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,461 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).

microsoft/vscode · 5,262 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,680 tokens

vscode copilot-instructions.md

Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).

microsoft/vscode · 2,750 tokens

deno copilot-instructions.md

Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.

denoland/deno · 3,077 tokens