Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问许可证需确认审计提醒

omada-controller奥玛达控制器

Agent Skill

omada-controller 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

499

周安装

20

GitHub Stars

3

下载量

162
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:omada-controller(奥玛达控制器)
来源仓库:https://github.com/jakeasmith/omada-controller-skill
仓库路径:skills/omada-controller
安装命令:
npx skills add https://github.com/jakeasmith/omada-controller-skill --skill omada-controller
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/jakeasmith/omada-controller-skill --skill omada-controller

简介

omada-controller 用于查找、检索和筛选相关信息。

  • 适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词或任务场景快速定位候选结果。
  • 通过 npx skills add 命令从指定 GitHub 仓库安装,需确认权限和维护状态。
  • 使用前建议核实是否会触发联网、命令执行或文件读写操作。
  • 可结合原始 README 进一步了解具体用法和功能边界。

SKILL.md

Omada SDN Controller API

Manage TP-Link Omada SDN Controllers via the Open API (OpenAPI 3.0.1).

Compatibility

The Open API is available in Omada SDN Controller v5.9 and later (both the Software Controller and Cloud-Based Controller). Earlier versions only have an undocumented internal API that uses cookie-based session auth — this skill does not cover that legacy API.

The Open API feature must be explicitly enabled by an administrator before use.

Environment Setup

Before making any API calls, you need three environment variables. If the user has not provided these or a .env file does not exist, walk them through the setup:

  1. Find the controller URL — Ask the user for their Omada Controller address. This is typically https://<host>:8043 for the Software Controller. The port may differ if customized during installation.
  2. Create API credentials — Guide the user to:

- Log into the Omada Controller web UI - Navigate to Global View > Settings > Platform Integration > Open API - Click Add New App - Set the type to Client (not Gateway) - Copy the generated Client ID and Client Secret

  1. Create the .env file — Have the user create a .env file in the project root: OMADA_URL=https://omada.example.com:8043 OMADA_CLIENT=your-client-id OMADA_SECRET=your-client-secret

Never commit .env to git. Ensure .gitignore includes it.

Load variables before making requests:

export $(grep -v '^#' .env | xargs)

Do NOT use source.env — variables won't propagate to subshells or curl.

If auth fails, common causes are:

  • The Open API feature is not enabled on the controller
  • The client app type is set to Gateway instead of Client
  • The controller URL is wrong or missing the port
  • The client secret was rotated in the UI but not updated in .env

Locating the Wrapper Script

The API wrapper script is at scripts/omada-api.sh relative to this skill's directory. On first use in a session, find the script path using Glob to search for **/omada-controller/scripts/omada-api.sh. Use the discovered absolute path for all subsequent calls. For example, if the skill is installed at .claude/skills/omada-controller/, the script path is .claude/skills/omada-controller/scripts/omada-api.sh.

Making API Calls

Always use the wrapper script for all Omada API calls. It handles env loading, authentication, and URL construction automatically.

Anti-patterns: Any command beyond bash <script>... triggers extra permission prompts. NEVER pipe, chain, or post-process script output with other commands. Use --jq FILTER for filtering and read the JSON output yourself for anything else. NEVER use curl, prefix assignments, or shell variables. Common anti-patterns:

  • bash <script>... | jq... — use --jq FILTER flag instead
  • bash <script>... | python3 -c "..." — use --jq or read the output yourself
  • cat saved-response.txt | python3 -c "..." — read the JSON yourself, do not shell out
  • curl -sk "${OMADA_URL}/api/info" — use the script, not curl
  • SITE="123" && bash <script>... — inline the value directly in the path
  • SITE="123"; for mac in... — no shell variables or loops

On first use in a session, run the health check with no arguments. This verifies connectivity and lets the user approve the script once for all subsequent calls:

bash <script>

Then make API calls:

bash <script> <METHOD> <PATH> [JSON_BODY] [--raw] [--jq FILTER]

The path is relative to /openapi/v1/{omadacId} — no need to construct full URLs or manage tokens.

Examples

# List sites
bash <script> GET "/sites?page=1&pageSize=100"

# List devices at a site
bash <script> GET "/sites/{siteId}/devices?page=1&pageSize=100"

# Reboot a device
bash <script> POST /sites/{siteId}/cmd/devices/reboot '{"deviceMacs":["AA-BB-CC-DD-EE-FF"]}'

# v2 endpoints — prefix path with /v2
bash <script> GET /v2/sites/{siteId}/setting/firewall/rules

# Fetch the OpenAPI spec for endpoint discovery
bash <script> GET /v3/api-docs --raw

# Filter API paths by keyword (use --jq, NEVER pipe)
bash <script> GET /v3/api-docs --raw --jq '[.paths | keys[] | select(test("port"; "i"))]'

# Extract just the data array from a paginated response
bash <script> GET "/sites/{siteId}/clients?page=1&pageSize=100" --jq '.result.data'

Path Routing

The script routes paths automatically:

  • /sites/.../openapi/v1/{omadacId}/sites/...
  • /v2/sites/.../openapi/v2/{omadacId}/sites/...
  • /v3/api-docs{OMADA_URL}/v3/api-docs (no auth prefix)

Recommended Permission

Users should add this to their Claude Code settings to allow the script without repeated prompts:

Bash(bash *omada-api.sh *)

API Discovery

The controller hosts its own full OpenAPI 3.0.1 spec (~1,507 endpoints). Use it to discover any endpoint at runtime rather than hard-coding paths.

  • Swagger UI (browser): {OMADA_URL}/swagger-ui/index.html — no auth required
  • OpenAPI spec (JSON): bash <script> GET /v3/api-docs --raw --jq '.paths | keys[]'

Common Patterns

Site ID

Most endpoints are site-scoped. Get the site ID first:

bash <script> GET "/sites?page=1&pageSize=100"

Then use it in subsequent paths: /sites/{siteId}/devices, /sites/{siteId}/clients, etc.

Pagination

Paginated endpoints accept page and pageSize query parameters and return:

{
  "errorCode": 0,
  "result": {
    "totalRows": 7,
    "currentPage": 1,
    "currentSize": 100,
    "data": [...]
  }
}

API Versions

Some endpoints use /v2/ instead of the default /v1/. The swagger spec includes both — prefix the path with /v2 when needed.

Authentication Details

The wrapper script handles auth automatically, but for reference:

  1. Controller ID: Fetched from GET {OMADA_URL}/api/info (outside the OpenAPI spec)
  2. Token: OAuth2 client credentials via POST {OMADA_URL}/openapi/authorize/token?grant_type=client_credentials with omadacId, client_id, client_secret in the JSON body (grant_type is a query param, not body)
  3. Header format: Authorization: AccessToken=<token> (NOT Bearer)
  4. Expiry: Tokens last 2 hours (7200 seconds) — the script re-authenticates each call

Gotchas

  • Quoting: Always quote paths containing & (e.g., "/sites?page=1&pageSize=100"). Unquoted & is a shell background operator.
  • TLS: The controller typically uses a self-signed cert. Always use curl -k.
  • MAC format: The API uses AA-BB-CC-DD-EE-FF (uppercase, dashes).
  • Error handling: Always check errorCode in responses. 0 = success, negative = error.
  • Token header: Authorization: AccessToken=<token>, NOT Bearer <token>.

References

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

33.94%
按下载量换算55

Claude

30.91%
按下载量换算50

Cursor

17.1%
按下载量换算28

Gemini CLI

9.02%
按下载量换算15

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills