Home / weiransoft / traemultiagentskill · GitHub

multi-agent-team SKILL.md skillA

multi-agent-team from weiransoft/traemultiagentskill is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add weiransoft/traemultiagentskill/multi-agent-team@git:20261016.f28da00

Pins git:20261016.f28da00 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
weiransoft/traemultiagentskill
Path
SKILL.md
Versions recorded
8
Latest version
git:20261016.f28da00, committed Oct 16, 2026
First version recorded
Sep 29, 2026
Size
93,201 bytes
Repository stars
46
License
MIT

基于任务类型动态调度到合适的智能体角色（架构师、产品经理、测试专家、独立开发者、UI 设计师）。支持多智能体协作、共识机制、完整项目生命周期管理、规范驱动开发、代码走读审查和项目理解能力。支持中英文双语。v2.4 新增 Karpathy 四大核心原则，v2.5 新增 Cybernetics 工程控制论增强，v2.6 新增 Ponytail 决策梯（少写多余代码）、Autonomous 自主迭代模式、Dynamic Workflows 6 大模式、插件热加载，v2.7 新增 UI/UX 巡检分析与视觉回归测试脚本，v2.7.1 修订 AI 诚实降级、真实语义匹配、双宿主清单同步与 v1 死代码清算，v2.9 新增既有系统理解能力（黑盒 Web 系统反向理解：Playwright 登录 + BFS 遍历 + DB/Redis 只读内省 + 10 节理解文档），v2.9.1 新增 SU 专家详说阶段（SFD：五专家分节撰写并装配 SYSTEM_FUNCTION_DOC.md），v2.9.2 修复 D1（understanding.json 的 meta.run_status…

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Multi-Agent Team Dispatcher (AI-Enhanced)
基于任务类型和上下文，自动调度到最合适的智能体角色（架构师、产品经理、测试专家、Solo Coder、UI 设计师）。

> **能力实现方式说明（v2.7.1 诚实标注）**：
> - **提示词层 AI（宿主 LLM 完成）**：任务语义理解、角色智能匹配、多角色共识决策、架构设计审查——由 Trae/Claude 宿主大模型直接执行，脚本仅提供候选清单与规则约束。
> - **脚本层确定性工具（Python 实现）**：代码地图扫描、UI/UX 巡检、视觉回归对比、决策梯合规检查、TFIDF/Hashing 文本相似度——为无 LLM 的独立进程，提供可复现的确定性结果。
> - **降级模式说明**：当脚本层需要语义相似度但无网络/模型时，自动降级到 TFIDF/Hashing 本地算法；SubAgent 调用在无真实 claude 命令时返回错误（`fallback_mode: error`），真实并行子代理由宿主 Task 机制实现。

**v2.7.1 修订（AI 诚实化 + 代码清算 + 双宿主同步）**:
- 🤥→✅ AI 诚实降级：`ai_assistant.py` / `ai_semantic_matcher.py` 移除全部模拟响应，无 AI 客户端时明确抛错并降级到确定性匹配，不再伪装 AI 输出；`claude_code_subagent_adapter.py` 移除 `_simulate_subagent_call`，无真实 subagent 时返回 `success=False`
- 🔢 真实语义匹配：`role_matcher.py` 从关键词重叠升级为本地 embedder（TFIDF/Hashing）向量余弦相似度
- 🛡️ embedder 三级降级链：`goal_orchestrator.py` SentenceTransformer → TFIDF → Hashing，无网络/无模型时自动降级
- 🎭 Claude Code 真实 SubAgent：`.claude/agents/` 新增 5 角色定义文件，替代脚本模拟
- 🔄 双宿主清单同步：`sync_manifests.py` CI 校验三份 manifest（skill-manifest.yaml / skills-index.json / claude-code-skill.json）版本与命名一致
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add weiransoft/traemultiagentskill/multi-agent-team@sha256:d121eb731f24695b

The line at the top of this page pins the label git:20261016.f28da00, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_5bgzqwbgekdk4tx5.svg)](https://markdownregistry.com/a/art_5bgzqwbgekdk4tx5)

2 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 8 versions of this file, between Sep 29, 2026 and Oct 16, 2026, about one every 2 days.

versioncommittedcommitsizeaudit
git:20261016.f28da00 latest2026-10-16 f28da00 93,201 BA view · diff
git:20261015.3796d492026-10-15 3796d49 86,966 BA view · diff
git:20261014.83bf76a2026-10-14 83bf76a 83,357 BA view · diff
git:20261012.bc967932026-10-12 bc96793 68,969 BA view · diff
git:20261011.e2babc82026-10-11 e2babc8 57,168 BA view · diff
git:20260930.775d0e62026-09-30 775d0e6 47,775 BA view · diff
git:20260930.f794ef42026-09-30 f794ef4 47,668 BA view · diff
git:20260929.34207ab2026-09-29 34207ab 46,199 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (93201 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

weiransoft/traemultiagentskill · 46 stars · license MIT · pushed 2026-10-18 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_5bgzqwbgekdk4tx5
GET https://markdownregistry.com/api/v1/resolve?ref=weiransoft/traemultiagentskill/multi-agent-team
GET https://markdownregistry.com/api/v1/blob/d121eb731f24695b0ee99a787dfc4bd7e5fa4a5d4bca3ac712b4c1db5e5edf34

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

Browse by kind, by grade A, or by owner.