Token导航 LogoToken导航

Agent Skill · 开发

ado-context(废话上下文)

ado-context 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。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。

ado-context 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描已通过,按安装量在开发类的 21,746 个 Skill 里排第 21,271。

总安装 192周安装 8GitHub Stars 公开资料未说明下载 64核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:ado-context(废话上下文)
来源仓库:https://github.com/lewisth/ado-skills
仓库路径:skills/ado-context
安装命令:
npx skills add https://github.com/lewisth/ado-skills --skill ado-context
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/lewisth/ado-skills --skill ado-context

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

ado-context 用于解析和管理 Azure DevOps 项目的共享上下文信息,如组织、项目、流程模板和仓库路径。适合在需要创建或更新工作项前初始化环境上下文的场景,支持与多个宿主工具集成。通过读取 .ado-skills.json 文件获取静态配置,动态计算迭代路径,需先运行一次设置步骤。

安装前请确认权限范围、维护状态,并注意是否会触发联网或文件读写操作。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/ado-context

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/lewisth/ado-skills --skill ado-context;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

ado-context 怎么安装?

在终端执行 npx skills add https://github.com/lewisth/ado-skills --skill ado-context。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

ado-context 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

ado-context 需要什么权限?

权限级别是接触敏感数据。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。

ado-context 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

ado-context 用的人多吗?

公开数据是累计安装 192 次,近一周安装 8 次,下载 64 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

ado-context 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理开发类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
ado-context

废话上下文 · 本页

192—接触敏感数据未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

ADO Context

Resolve the shared Azure DevOps context used by every other ADO skill:

  • ORG — organisation URL (e.g. https://dev.azure.com/contoso)
  • ORG_NAME — short org name (e.g. contoso)
  • PROJECT — project name
  • PROCESS — process template name (Agile / Scrum / CMMI)
  • REPOSITORY_URL — canonical URL of the repo the agent should work in
  • AREA_PATH — where work items are filed
  • ITERATION_PATH — which sprint/iteration they land in

All but ITERATION_PATH are static per repo — they don't move. Resolve them once in a setup step that writes .ado-skills.json, then just read the file on every subsequent invocation. The iteration path is always computed fresh from today's date.

Run this before any skill that creates or updates work items (to-feature, to-pbis, …). Export the values so downstream steps can reference them without re-resolving.

Shell environment

These skills assume a Bash-compatible shell (Linux, macOS, Git Bash, WSL). If the workspace shell is PowerShell, use the PowerShell equivalents shown alongside each Bash block. Every shell invocation in an agent context is a fresh process — environment variables do not persist between calls. Set all required variables (including $env:AZURE_DEVOPS_PAT) at the top of every script block that needs them.

Authentication

All REST calls use a Personal Access Token (PAT). The PAT must be set in the environment as AZURE_DEVOPS_PAT with at least Read scope on Project and Work Items (and Read & Write for skills that create items).

Build the auth header once at the start of every skill that calls this:

Bash:

AUTH=$(printf ':%s' "$AZURE_DEVOPS_PAT" | base64)
# Usage: -H "Authorization: Basic $AUTH"

PowerShell:

$AUTH = [System.Convert]::ToBase64String([System.Text.Encoding]::ASCII.GetBytes(":$env:AZURE_DEVOPS_PAT"))
# Usage: -Headers @{ Authorization = "Basic $AUTH" }

If AZURE_DEVOPS_PAT is unset or empty, stop and tell the user to set it before continuing. In agent contexts, set it at the top of every script block — it will not carry over from a previous invocation.

Config file

.ado-skills.json lives at the repo root and is committed so the whole team shares the same defaults. All fields are required — the skill will fail if any are missing:

{
  "organizationUrl": "https://dev.azure.com/contoso",
  "project": "My Project",
  "process": "Scrum",
  "repositoryUrl": "https://dev.azure.com/contoso/My Project/_git/my-repo",
  "areaPath": "My Project\\Squad A",
  "team": "Engineering"
}

Backslashes must be escaped in JSON. iterationPath is not stored — it's calculated on every run. team is required — do not leave it out or default it; ask the user for the correct team name if unknown.

Process

0. Set up auth

Bash:

if [ -z "$AZURE_DEVOPS_PAT" ]; then
  echo "AZURE_DEVOPS_PAT is not set. Please create a PAT in Azure DevOps and set it."
  exit 1
fi
AUTH=$(printf ':%s' "$AZURE_DEVOPS_PAT" | base64)

PowerShell:

if (-not $env:AZURE_DEVOPS_PAT) {
  Write-Error "AZURE_DEVOPS_PAT is not set. Please create a PAT in Azure DevOps and set it."
  exit 1
}
$AUTH = [System.Convert]::ToBase64String([System.Text.Encoding]::ASCII.GetBytes(":$env:AZURE_DEVOPS_PAT"))

1. Load .ado-skills.json if it exists

Bash:

CONFIG_FILE=".ado-skills.json"

if [ -f "$CONFIG_FILE" ]; then
  export ORG=$(jq -r '.organizationUrl // empty'      "$CONFIG_FILE")
  export PROJECT=$(jq -r '.project // empty'          "$CONFIG_FILE")
  export PROCESS=$(jq -r '.process // empty'          "$CONFIG_FILE")
  export REPOSITORY_URL=$(jq -r '.repositoryUrl // empty' "$CONFIG_FILE")
  export AREA_PATH=$(jq -r '.areaPath // empty'       "$CONFIG_FILE")
  export TEAM=$(jq -r '.team // empty'                "$CONFIG_FILE")
  export ORG_NAME=$(basename "$ORG")
fi

PowerShell:

$CONFIG_FILE = ".ado-skills.json"

if (Test-Path $CONFIG_FILE) {
  $config = Get-Content $CONFIG_FILE | ConvertFrom-Json
  $ORG            = $config.organizationUrl
  $PROJECT        = $config.project
  $PROCESS        = $config.process
  $REPOSITORY_URL = $config.repositoryUrl
  $AREA_PATH      = $config.areaPath
  $TEAM           = $config.team
  $ORG_NAME       = $ORG.TrimEnd('/').Split('/')[-1]
}

After loading, validate that every required field is non-empty, including team. If any field is missing, stop and run step 2 to resolve the missing values — even if all other fields are present. Do not silently default team to {project} Team; ask the user for the correct value.

2. First-time setup (only when config is missing or incomplete)

2a. Detect org, project, and repository URL from git remote

Bash:

REMOTE=$(git remote get-url origin)

PowerShell:

$REMOTE = git remote get-url origin

Match one of these shapes and extract the parts:

  • HTTPS: https://dev.azure.com/{org}/{project}/_git/{repo}
  • SSH: git@ssh.dev.azure.com:v3/{org}/{project}/{repo}
  • Legacy: https://{org}.visualstudio.com/{project}/_git/{repo}

Normalise the repository URL to the canonical HTTPS form so agents always check out from the same place, regardless of who originally cloned via SSH:

Bash:

ORG_NAME="contoso"
PROJECT="My Project"
REPO="my-repo"

ORG="https://dev.azure.com/$ORG_NAME"
REPOSITORY_URL="$ORG/$PROJECT/_git/$REPO"

PowerShell:

$ORG_NAME = "contoso"
$PROJECT = "My Project"
$REPO = "my-repo"

$ORG = "https://dev.azure.com/$ORG_NAME"
$REPOSITORY_URL = "$ORG/$PROJECT/_git/$REPO"

Confirm the derived values with the user before persisting — especially if the remote is a fork or mirror.

2b. Detect the process template

URL-encode the project name (spaces become %20) and call the Projects API:

Bash:

PROJECT_ENCODED=$(printf '%s' "$PROJECT" | jq -sRr @uri)

PROCESS=$(curl -s \
  -H "Authorization: Basic $AUTH" \
  "https://dev.azure.com/$ORG_NAME/_apis/projects/$PROJECT_ENCODED?includeCapabilities=true&api-version=7.1" \
  | jq -r '.capabilities.processTemplate.templateName')

PowerShell:

$PROJECT_ENCODED = [Uri]::EscapeDataString($PROJECT)

$resp = Invoke-RestMethod `
  -Uri "https://dev.azure.com/$ORG_NAME/_apis/projects/$PROJECT_ENCODED?includeCapabilities=true&api-version=7.1" `
  -Headers @{ Authorization = "Basic $AUTH" }
$PROCESS = $resp.capabilities.processTemplate.templateName

This picks the right story type downstream (User Story / Product Backlog Item / Requirement).

2c. Pick an area path

Fetch the area tree and flatten it to a list of paths for the user to choose from:

Bash:

curl -s \
  -H "Authorization: Basic $AUTH" \
  "https://dev.azure.com/$ORG_NAME/$PROJECT_ENCODED/_apis/wit/classificationnodes/areas?\$depth=10&api-version=7.1" \
  | jq -r 'recurse(.children[]?) | .path'

PowerShell:

$areas = Invoke-RestMethod `
  -Uri "https://dev.azure.com/$ORG_NAME/$PROJECT_ENCODED/_apis/wit/classificationnodes/areas?`$depth=10&api-version=7.1" `
  -Headers @{ Authorization = "Basic $AUTH" }

function Get-AreaPaths($node) {
  $node.path
  if ($node.children) { $node.children | ForEach-Object { Get-AreaPaths $_ } }
}
Get-AreaPaths $areas

Present the list and prompt the user to pick one. If the user has no preference, fall back to the project root ($PROJECT) — that's ADO's own default.

2d. Determine the team name

The iteration endpoint is team-scoped. Ask the user for the exact team name — do not guess or default. The team name must exactly match an existing team in the project:

Bash:

# Ask the user: "What is the name of your ADO team?"
TEAM="Engineering"   # example — use the value the user provides

PowerShell:

# Ask the user: "What is the name of your ADO team?"
$TEAM = "Engineering"   # example — use the value the user provides
2e. Write .ado-skills.json

Bash:

jq -n \
  --arg org     "$ORG" \
  --arg proj    "$PROJECT" \
  --arg proc    "$PROCESS" \
  --arg repo    "$REPOSITORY_URL" \
  --arg area    "$AREA_PATH" \
  --arg team    "$TEAM" \
  '{
    organizationUrl: $org,
    project:         $proj,
    process:         $proc,
    repositoryUrl:   $repo,
    areaPath:        $area,
    team:            $team
  }' > "$CONFIG_FILE"

PowerShell:

@{
  organizationUrl = $ORG
  project         = $PROJECT
  process         = $PROCESS
  repositoryUrl   = $REPOSITORY_URL
  areaPath        = $AREA_PATH
  team            = $TEAM
} | ConvertTo-Json | Set-Content $CONFIG_FILE

Tell the user to commit .ado-skills.json so teammates and CI agents share the same context. Export the variables so the rest of this run can use them.

3. Resolve iteration path (from today's date)

Fetch the current iteration for the team. ADO determines which iteration contains today — no date math needed:

Bash:

TEAM_ENCODED=$(printf '%s' "$TEAM" | jq -sRr @uri)

ITERATION_PATH=$(curl -s \
  -H "Authorization: Basic $AUTH" \
  "https://dev.azure.com/$ORG_NAME/$PROJECT_ENCODED/$TEAM_ENCODED/_apis/work/teamsettings/iterations?\$timeframe=current&api-version=7.1" \
  | jq -r '.value[0].path // empty')

PowerShell:

$TEAM_ENCODED = [Uri]::EscapeDataString($TEAM)

$iterResp = Invoke-RestMethod `
  -Uri "https://dev.azure.com/$ORG_NAME/$PROJECT_ENCODED/$TEAM_ENCODED/_apis/work/teamsettings/iterations?`$timeframe=current&api-version=7.1" `
  -Headers @{ Authorization = "Basic $AUTH" }
$ITERATION_PATH = $iterResp.value[0].path

If nothing comes back — no sprint configured for today, or a gap between sprints — fall back to the project default iteration and warn the user:

Bash:

if [ -z "$ITERATION_PATH" ]; then
  echo "No current iteration found — falling back to project default."
  ITERATION_PATH="$PROJECT"
fi

export ITERATION_PATH

PowerShell:

if (-not $ITERATION_PATH) {
  Write-Warning "No current iteration found — falling back to project default."
  $ITERATION_PATH = $PROJECT
}

4. Report

Print the resolved context so the user can sanity-check before work items get created:

ORG:            https://dev.azure.com/contoso
PROJECT:        My Project
PROCESS:        Scrum
REPOSITORY_URL: https://dev.azure.com/contoso/My Project/_git/my-repo
AREA_PATH:      My Project\Squad A
ITERATION_PATH: My Project\Sprint 42
TEAM:           Engineering

平台分布

Codex

36.27%
按下载量换算23

Claude

29.72%
按下载量换算19

Cursor

15.84%
按下载量换算10

Gemini CLI

9.1%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

接触敏感数据

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills