Agent skill

Fkteams

by wsshow in wsshow/feikong-teams

fkteams 多智能体协作 AI 助手的完整使用指南。当用户需要启动 fkteams、切换工作模式(团队/深度/讨论/自定义)、 通过命令行或管道执行查询、管理单个智能体(coder/researcher/fkteamshelper/analyst/remote/generalist)、管理会话历史(保存/加载/导出/恢复)、…

MITAuto-check: warnings

Install Fkteams

The automated check flagged lines worth reading first. See the safety section below.

skills CLI
$ npx skills add wsshow/feikong-teams --skill fkteams -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install wsshow/feikong-teams fkteams --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/wsshow/feikong-teams.git skills-src && mkdir -p .claude/skills && cp -r skills-src/docs/skills/fkteams .claude/skills/fkteams && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
fkteams
GitHub stars
153
Token cost
~2.1k tokens
SKILL.md length
273 words
Files
1
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

fkteams 多智能体协作 AI 助手的完整使用指南。当用户需要启动 fkteams、切换工作模式(团队/深度/讨论/自定义)、 通过命令行或管道执行查询、管理单个智能体(coder/researcher/fkteamshelper/analyst/remote/generalist)、管理会话历史(保存/加载/导出/恢复)、…

  • SKILL.md covers 一、快速入门, 二、启动与运行模式, 三、交互模式内置命令 and 四、单智能体模式(agent 子命令), plus 6 more sections
  • Calls git, curl and ssh; reaches api.openai.com

What it does

Fkteams is an agent skill from wsshow/feikong-teams. fkteams 多智能体协作 AI 助手的完整使用指南。当用户需要启动 fkteams、切换工作模式(团队/深度/讨论/自定义)、 通过命令行或管道执行查询、管理单个智能体(coder/researcher/fkteamshelper/analyst/remote/generalist)、管理会话历史(保存/加载/导出/恢复)、 管理模型配置(添加/切换/删除/登录服务商)、管理本地技能(列出/搜索/安装/移除)、初始化或修改配置文件, 以及了解 fkteams 的任意命令行用法时,请使用此 skill。

Its SKILL.md is about 2.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts. Compatibility notes: 需要 fkteams 二进制文件,首次使用前需运行 fkteams generate config 生成配置文件。

It works with Model Context Protocol and OpenAI. The repository describes itself as: 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。 The licence is MIT.

Example prompts

  • “Use the fkteams skill to fkteam 多智能体协作 AI 助手的完整使用指南。当用户需要启动 fkteams、切换工作模式(团队/深度/讨论/自定义)、 通过命令行或管道执行查询、管理单个智能体(coder/researcher/fkteamshelper/analyst…”
  • “/fkteams”

Requirements

  • Python 3
  • Compatibility (from SKILL.md): 需要 fkteams 二进制文件,首次使用前需运行 fkteams generate config 生成配置文件。

What it can do on your machine

Read from SKILL.md and the folder at commit 8b9fcc3. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • curl
    • ssh

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • api.openai.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

  • Compatibility

    需要 fkteams 二进制文件,首次使用前需运行 fkteams generate config 生成配置文件。

    From compatibility in the SKILL.md frontmatter.

Context cost

Fkteams loads about 2.1k tokens when it runs. Until then it costs about 66 tokens; SKILL.md has 273 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~66
When it runs · the whole SKILL.md, loaded when a task matches
~2.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: warnings

The automated check found patterns that need a careful read before installing.

  • WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:330
    password = "pass", known_hosts_file = "~/.ssh/known_hosts" }

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from wsshow/feikong-teams at commit 8b9fcc3, republished under its MIT licence (© wsshow). 273 words, ~2,132 tokens.

Download SKILL.mdSave it as .claude/skills/fkteams/SKILL.md (or your agent's skills folder).
name
fkteams
description
fkteams 多智能体协作 AI 助手的完整使用指南。当用户需要启动 fkteams、切换工作模式(团队/深度/讨论/自定义)、 通过命令行或管道执行查询、管理单个智能体(coder/researcher/fkteams_helper/analyst/remote/generalist)、管理会话历史(保存/加载/导出/恢复)、 管理模型配置(添加/切换/删除/登录服务商)、管理本地技能(列出/搜索/安装/移除)、初始化或修改配置文件, 以及了解 fkteams 的任意命令行用法时,请使用此 skill。
compatibility
需要 fkteams 二进制文件,首次使用前需运行 fkteams generate config 生成配置文件。
metadata.author
fkteams
metadata.version
1.0

fkteams 命令行完整指南

一、快速入门

bash
# 第一步:生成配置文件
fkteams generate config

# 第二步:编辑配置,添加模型(必填)
# 编辑 ~/.fkteams/config/config.toml,填写 api_key 和 model

# 第三步(可选):安装运行时依赖
fkteams init --all   # 安装 uv(Python)和 bun(JavaScript)

# 第四步:启动
fkteams web          # Web 界面(推荐)
fkteams              # CLI 交互模式

二、启动与运行模式

模式对照表
模式命令适用场景
Web 界面fkteams web日常使用,需要历史记录和可视化界面
CLI 交互fkteams服务器/终端环境
直接查询fkteams -q "..."脚本集成、自动化
纯 API 服务fkteams serve作为后端独立部署
Web 界面模式
bash
fkteams web
# 启动后访问 http://localhost:23456
CLI 交互模式
bash
fkteams              # 默认:团队模式(coordinator 协调多智能体)
fkteams -m deep      # 深度分析模式
fkteams -m group     # 多智能体讨论模式(圆桌)
fkteams --temporary  # 临时会话,不保存历史
直接查询模式(非交互,执行后退出)
bash
fkteams -q "帮我审查 main.go"
fkteams -m deep -q "深度分析这个架构"
fkteams -q "生成一份报告"             # 默认保存历史
fkteams -q "临时问答" --temporary     # 不保存历史
管道输入模式

管道有内容时自动进入非交互模式。

bash
echo "解释一下 Go 的 context 包" | fkteams
cat main.go | fkteams -q "审查以下代码:"
git diff HEAD~1 | fkteams -q "审查这次提交"
curl -s https://example.com/api | fkteams -q "解析这个 API 响应"

规则:同时提供 -q 时,查询内容为 -q 文本 + 换行 + 管道内容;管道为空且无 -q 时报错。

纯 API 服务模式
bash
fkteams serve
fkteams serve --host 0.0.0.0 --port 8080

与 web 提供相同的 API,但无前端页面。支持端点:GET /v1/models、POST /v1/chat/completions。

恢复历史会话
bash
fkteams -r "20260302_091249"              # 交互模式恢复
fkteams -r "20260302_091249" -q "继续上次的分析"  # 恢复后直接查询
全局参数
参数简写说明
--mode-m工作模式:team(默认)/ deep / group
--query-q直接查询模式,执行后退出
--resume-r恢复指定会话 ID
--temporary--temp临时会话,不保存历史
--approve自动批准工具调用:all / command / file / git / dispatch(逗号分隔)

三、交互模式内置命令

命令说明
quit / q退出程序
help显示帮助
list_agents列出所有可用智能体
@智能体名 [查询]切换到指定智能体并可选执行查询
switch_work_mode切换工作模式
save_chat_history保存当前会话
list_chat_history列出所有历史会话
load_chat_history选择并加载历史会话
clear_chat_history清空当前会话(不删除文件)
save_chat_history_to_markdown导出为 Markdown
save_chat_history_to_html导出为 HTML
list_schedule列出所有定时任务
cancel_schedule取消定时任务
delete_schedule删除定时任务
list_memory列出长期记忆条目
delete_memory删除记忆条目
clear_memory清空所有长期记忆

四、单智能体模式(agent 子命令)

智能体目录
名称配置项角色
coordinator[[agents.items]] id = "coordinator"协调者
coder[[agents.items]] id = "coder"软件工程师,代码实现、调试、重构
researcher[[agents.items]] id = "researcher"DuckDuckGo 网络搜索
fkteams_helper[[agents.items]] id = "fkteams_helper"fkteams 安装、配置、扩展与故障排查答疑
analyst[[agents.items]] id = "analyst"数据分析(Excel、Python、文档)
remote[[agents.items]] id = "remote" + ssh = { ... }SSH 远程服务器访问
generalist[[agents.items]] id = "generalist"通用执行助手,支持多工具任务
agent 命令用法
bash
# 列出所有可用智能体
fkteams agent list

# 交互模式(进入对话)
fkteams agent -n coder
fkteams agent --name analyst

# 直接查询(一次性,执行后退出)
fkteams agent -n researcher -q "搜索最新的 Go 语言新闻"
fkteams agent -n coder -q "解释这个函数的作用"
fkteams agent -n fkteams_helper -q "如何配置 MCP?"

# 配合管道
cat error.log | fkteams agent -n coder -q "分析这个错误日志"
git diff HEAD~1 | fkteams agent -n coder -q "审查这次提交"
cat data.csv | fkteams agent -n analyst -q "计算基本统计数据"

# JSON 原始事件输出(用于程序化处理)
fkteams agent -n researcher -q "搜索 AI 新闻" --format json

# 自动批准工具调用
fkteams agent -n shell -q "清理临时文件" --approve all
fkteams agent -n coder -q "重构 main.go" --approve file,command

# 临时运行,不保存历史
fkteams agent -n researcher -q "搜索 AI 新闻" --temporary
agent 子命令参数
参数简写说明
list列出所有可用智能体
--name-n智能体名称(必填,与 list 互斥)
--query-q直接查询模式
--temporary--temp临时会话,不保存历史
--format输出格式:default(格式化)或 json(原始事件)
--approve自动批准:all / command / file / git / dispatch

交互模式下,输入 @ 符号后自动显示智能体列表供选择。


五、会话管理(session 子命令)

会话文件保存在 ~/.fkteams/sessions/,以时间戳命名(如 20260302_091249)。

bash
# 列出所有历史会话
fkteams session list

# 会话默认保存(退出时写入文件)
fkteams
fkteams -q "你的问题"

# 恢复历史会话
fkteams -r "20260302_091249"
fkteams -r "20260302_091249" -q "继续上次的问题"

交互模式内:save_chat_history / load_chat_history / clear_chat_history / save_chat_history_to_markdown / save_chat_history_to_html


六、模型管理(model 子命令)

bash
# 列出已配置的模型
fkteams model ls       # 或 fkteams model list

# 查询服务商的可用模型
fkteams model lr --name deepseek
fkteams model lr --provider openai

# 切换默认对话模型(交互式选择)
fkteams model sw
# 指定配置名
fkteams model sw --name deepseek
# 切换到指定模型
fkteams model sw --name deepseek --model deepseek-reasoner

# 移除模型配置(交互式选择)
fkteams model rm
fkteams model rm --name old-config
登录服务商(写入 config.toml,无需手动编辑)
bash
fkteams login openai     --api-key sk-...
fkteams login deepseek   --api-key sk-...
fkteams login claude     --api-key sk-ant-...
fkteams login gemini     --api-key AIza...
fkteams login qwen       --api-key sk-...
fkteams login ollama                          # 无需 API Key
fkteams login ark        --api-key ...
fkteams login openrouter --api-key sk-or-...
fkteams login copilot                         # OAuth 设备码流程
fkteams login copilot --import                # 从 VS Code 已保存的 token 导入
fkteams login custom --base-url https://my-proxy.example.com/v1 --api-key sk-...

# 通用可选参数
fkteams login openai --api-key sk-... --model gpt-4o --name my-openai
# --name 指定模型 ID;首次登录会自动设为默认对话模型

# 退出登录
fkteams logout

七、技能管理(skill 子命令)

技能目录:~/.fkteams/skills/<技能名>/,每个技能必须包含 SKILL.md。

bash
# 列出本地已安装的技能
fkteams skill list    # 或 fkteams skill ls

# 搜索技能市场(默认后端:SkillHub)
fkteams skill search <关键词>
fkteams skill search ffmpeg --page 2 --size 20
fkteams skill search ffmpeg --provider SkillHub

# 安装技能
fkteams skill install <技能slug>
fkteams skill install video-frames --version 1.0.0
fkteams skill install video-frames --provider SkillHub

# 移除技能
fkteams skill remove <技能slug>

八、配置与初始化(generate / init 子命令)

bash
# 生成示例配置文件(首次使用必须执行)
fkteams generate config
# 生成路径:~/.fkteams/config/config.toml

# 生成 OpenAI 兼容 API 密钥
fkteams generate apikey

# 初始化运行时依赖
fkteams init           # 交互式选择
fkteams init --all     # 安装全部(uv + bun)
fkteams init --env uv  # 仅安装 uv(Python 脚本工具)
fkteams init --env bun # 仅安装 bun(JS 脚本工具)
fkteams init --mirror  # 生成镜像源配置(国内加速)
关键配置项
toml
# 模型(必填)
[[models]]
name     = "default"
provider = "openai"
base_url = "https://api.openai.com/v1"
api_key  = "sk-..."
model    = "gpt-4o"

# 智能体目录
[[agents.items]]
id = "researcher"
name = "研究员"
description = "网络研究员"
tools = ["search", "fetch"]
enabled = true

[[agents.items]]
id = "remote-prod"
name = "生产服务器"
description = "通过 SSH 管理生产服务器"
tools = ["ssh"]
ssh = { host = "ip:port", username = "user", password = "pass", known_hosts_file = "~/.ssh/known_hosts" }
enabled = true

# 长期记忆
[memory]
enabled = true

# Web 服务器
[server]
port = 23456

# OpenAI 兼容 API(需要先 fkteams generate apikey)
[openai_api]
api_keys = ["sk-fkteams-your-secret"]
环境变量
变量默认值说明
FEIKONG_APP_DIR~/.fkteams应用数据目录
FEIKONG_PROXY_URL—代理地址(唯一的代理配置方式)
FEIKONG_MAX_ITERATIONS60智能体最大迭代次数(0/-1 不限制)

九、其他子命令

bash
# 检查并更新 fkteams 到最新版本
fkteams update

# 列出所有可用工具
fkteams tool list

# 显示版本
fkteams --version

十、常见问题(Gotchas)

  • 会话默认保存:退出时会自动写入历史;临时任务可加 --temporary,额外导出可在交互中执行 save_chat_history_to_html 或 save_chat_history_to_markdown。
  • 管道触发非交互模式:哪怕管道为空也会触发,空管道 + 无 -q 会报错。
  • fkteams web vs fkteams serve:web 包含前端页面;serve 仅提供 API,无界面。
  • 首次使用需先生成配置:运行 fkteams generate config 后再启动,否则报错。
  • 可选智能体需先在配置中启用:调用未启用的智能体会返回错误。
  • 代理只能通过环境变量设置:config.toml 中没有代理配置项,只用 FEIKONG_PROXY_URL。
  • model sw 只切换默认,不删除配置:要删除用 model rm。
  • 技能安装后需重启 fkteams 才生效:技能在启动时加载,热安装不生效。
  • [openai_api] 不配置 api_keys 时所有 API 请求返回 401。

© wsshow, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in docs/skills/fkteams of wsshow/feikong-teams.

Open the folder on GitHubat commit 8b9fcc3

Compare with similar skills

Fkteams next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Fkteams compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Fkteams this skillwsshow/feikong-teams153—~2.1kAutomated safety check: WarnMIT
Annotate Paper54yyyu/zotero-mcp5.3k—~1.5kAutomated safety check: PassMIT
Research ReviewGRIND-Lab-Core/night_owl_research_agent1065 repos~1.1kAutomated safety check: NotesNone
Research RefinezjYao36/Auto-Research-Refine1286 repos~6.9kAutomated safety check: NotesNone
CLI Anything ZoteroPiaoyangGuohai1/cli-anything-zotero138—~2.6kAutomated safety check: PassApache-2.0
Deep Research MCP Guidepminervini/deep-research-mcp114—~5.8kAutomated safety check: PassMIT

Similar skills

  • Annotate Paper

    54yyyu/zotero-mcp

    Read the open paper and write study annotations into its PDF with zotero-cli - a context box on the title, a four-part summary on the abstract, role-coded abstract highlights, one box per figure…

    5.3k GitHub stars~1.5k tokensUpdated 3 days ago
    Research & ScienceAuto-check passed
  • Research Review

    GRIND-Lab-Core/night_owl_research_agent

    Get a deep critical review of research idea from GPT via Codex MCP.

    106 GitHub starsUsed in 5 repos~1.1k tokens
    Research & ScienceAuto-check: notes
  • Research Refine

    zjYao36/Auto-Research-Refine

    Turns a vague research direction into a focused, problem-anchored method plan through up to five review rounds with a second model.

    128 GitHub starsUsed in 6 repos~6.9k tokens
    Research & ScienceAuto-check: notes
  • CLI Anything Zotero

    PiaoyangGuohai1/cli-anything-zotero

    Full-featured CLI for Zotero reference management. An agent skill from PiaoyangGuohai1/cli-anything-zotero.

    138 GitHub stars~2.6k tokensUpdated 2 mo ago
    Research & ScienceAuto-check passed
  • Deep Research MCP Guide

    pminervini/deep-research-mcp

    Explains how to run, integrate and debug the deep-research-mcp project through its CLI, Python API or MCP server, with OpenAI, Gemini and DR-Tulu backends.

    114 GitHub stars~5.8k tokensUpdated 13 days ago
    Research & ScienceAuto-check passed
  • Research Review

    wanshuiyin/Auto-claude-code-research-in-sleep

    Get a deep critical review of research from an external reviewer backend (Codex or manual).

    17k GitHub stars~3.1k tokensUpdated 4 days ago
    Research & ScienceAuto-check: notes

Questions about Fkteams

What does Fkteams do?

fkteams 多智能体协作 AI 助手的完整使用指南。当用户需要启动 fkteams、切换工作模式(团队/深度/讨论/自定义)、 通过命令行或管道执行查询、管理单个智能体(coder/researcher/fkteamshelper/analyst/remote/generalist)、管理会话历史(保存/加载/导出/恢复)、…. Fkteams is an agent skill from wsshow/feikong-teams.

How do I install Fkteams in Claude Code?

Run `npx skills add wsshow/feikong-teams --skill fkteams -a claude-code`. Or copy the skill folder (docs/skills/fkteams in wsshow/feikong-teams) into .claude/skills/fkteams in your project. Claude Code loads it when a task matches its description.

How do I install Fkteams in Codex?

Run `npx skills add wsshow/feikong-teams --skill fkteams -a codex`. Or copy the skill folder (docs/skills/fkteams in wsshow/feikong-teams) into .agents/skills/fkteams in your project. Codex loads it when a task matches its description.

Can I use Fkteams in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add wsshow/feikong-teams --skill fkteams -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/fkteams, .gemini/skills/fkteams, .github/skills/fkteams and .opencode/skills/fkteams in your project.

What does Fkteams need to run?

Going by SKILL.md and its folder, Fkteams needs the command-line tools its instructions call (git, curl and ssh). Our summary lists: Python 3. Compatibility (from SKILL.md): 需要 fkteams 二进制文件,首次使用前需运行 fkteams generate config 生成配置文件。.

Does Fkteams access the network?

SKILL.md names 1 domain. In commands or code: api.openai.com; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.

Is Fkteams safe to install?

Our automated static check of SKILL.md flagged 1 warning(s): mentions a credentials file (ssh keys, cloud or package-manager tokens). Read the flagged lines before installing; the check is not a guarantee either way.

What licence does Fkteams use?

Fkteams is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Fkteams use?

About 2.1k tokens (SKILL.md is roughly 8.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Fkteams?

Skills that share tags, products or a category with Fkteams: Annotate Paper (54yyyu/zotero-mcp, 5.3k stars), Research Review (GRIND-Lab-Core/night_owl_research_agent, 106 stars), Research Refine (zjYao36/Auto-Research-Refine, 128 stars) and CLI Anything Zotero (PiaoyangGuohai1/cli-anything-zotero, 138 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Fkteams?

wsshow (a GitHub user) maintains it in wsshow/feikong-teams, which has 153 GitHub stars. The repository was last updated on September 5, 2026.

Source: wsshow/feikong-teams on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.