wsshow/feikong-teams — 一个基于多智能体协作的 AI 助手，支持命令行和Web界面，提供团队模式、自定义会议模式和多智能体讨论模式（圆桌会议模式）三种工作方式，通过多个专业智能体协同工作来完成复杂的编程和系统任务。

Description and use

**Brief Summary**
feikong‑teams is an open‑source AI assistant built on multi‑agent collaboration, offering a CLI and a web UI with three work modes—team mode, custom meeting mode, and round‑table discussion mode. By orchestrating specialized agents, it can tackle complex programming and system‑integration tasks that would be cumbersome for a single prompt‑based model.

**Value**
The platform turns ad‑hoc prompts and isolated tools into repeatable, orchestrated workflows, enabling teams to coordinate multiple agents, embed tool‑use pipelines, and maintain consistent agent memory. This reduces the overhead of manually chaining LLM calls, improves reliability for complex tasks, and provides a structured environment for collaborative AI‑driven development.

**Practical Adoption Path**

1. **Evaluation** – Clone the repository, run the provided Docker/CLI starter, and experiment with the built‑in team and meeting modes on a small code‑base.
2. **Integration** – Use the exposed Go SDK or REST API to embed feikong‑teams into existing CI/CD pipelines or internal dev‑tools, customizing agents for your domain‑specific tasks.
3. **Scaling** – Deploy the web UI behind an internal gateway, configure persistent storage for agent memory, and add custom tool plugins (e.g., code linters, cloud APIs) to extend the workflow.

**Production Readiness**
With recent commits (as of 2026‑06‑26), 121 stars, 28 forks, and active issue handling, the project shows strong community momentum. Its Go codebase, clear CLI/API surface, and modular architecture make it straightforward to containerize and monitor in production. While a final review of licensing, security hardening, and maintainer commitment is advisable, the overall signals indicate that feikong‑teams is mature enough for pilot deployments and can be hardened for enterprise use.

Problem

wsshow/feikong-teams helps turn isolated prompts and tools into repeatable agent workflows.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• coordinate multi-agent workflows
• add tool-use pipelines
• standardize agent memory

Evidence

一个基于多智能体协作的 AI 助手，支持命令行和Web界面，提供团队模式、自定义会议模式和多智能体讨论模式（圆桌会议模式）三种工作方式，通过多个专业智能体协同工作来完成复杂的编程和系统任务。

Source