Back to timeline

Thu, October 1511:15ToolsAI codingChinese modelsAgentsAI coding guide

agency-agents-zh Chinese agent library released: 267 plug-and-play expert roles

Decision Brief

What changedA Chinese community fork of agency-agents has appeared on GitHub, offering 267 plug-and-play AI expert roles and support for 20 mainstream AI coding tools.
Why it mattersFor teams using Claude Code, Cursor, Copilot and similar tools, this role library upgrades prompts from generic templates to role definitions with personas, workflows and deliverables, and adds coverage for China-market platforms like Xiaohongshu, Douyin, Feishu and DingTalk—saving the effort of tuning prompts yourself.
Who should careAI coding tool users
Affected stackClaude CodeOpenAIGeminiQwen
Builder actionUpdate Claude Code to the latest version
Source confidenceHigh · Official release / blog / repo

This is the Chinese community edition of agency-agents, containing 267 plug-and-play AI expert roles spanning company-management posts such as CEO, CTO and CFO, plus 20 departments including engineering, design, marketing, finance, legal and supply chain. Beyond fully translating the upstream 213 roles, it adds 64 original China-market agents covering operations on Xiaohongshu, Douyin, WeChat, Bilibili, Feishu and DingTalk, as well as verticals like cross-border e-commerce, government ToG, healthcare compliance, Qt industrial HMI and mechanical design. Each agent has its own persona, professional workflow and deliverable rather than being a generic prompt template. For installation, the repo provides a one-command install script supporting 20 tools including Claude Code, GitHub Copilot, Cursor, Windsurf, Gemini CLI, Qwen Code, Codex CLI and OpenClaw. Claude Code and GitHub Copilot install directly; other tools require running a conversion script first. OpenClaw users get three files—SOUL.md, AGENTS.md and IDENTITY.md—with built-in multi-agent collaboration. A companion orchestrator, agency-orchestrator, installs globally via npm and lets multiple experts auto-team up on a DAG with a single command, supporting no-code orchestration, resumable runs and 10 LLM integrations, 7 of which need no API key. For developer teams looking to land multi-agent workflows in concrete business scenarios, this role library plus orchestrator combo lowers the cost of moving from single prompts to multi-role collaboration.

Summary basis: full article readCompiled from the source scope noted above; the original remains authoritative.

Sources

Related intel

留言

登入后即可留言,和其他 builder 交换实测心得。

还没有留言,抢头香。