跳到主要内容
知仓学习社ZHICANG

live-weather-briefing-skill

>-

读凭据改身份文件执行命令读文件写文件联网严重 0 · 高危 22FrancyJGLisboa/agent-skills-platform

它会碰到什么

扫了多少22 个文本文件,159 KB
它会碰到什么读凭据改身份文件执行命令读文件写文件联网
命中总数55 处
命中统计严重 0 · 高 22 · 中 26 · 低 7
逐条看命中(22 条严重或高危)
  • install.ps1:30cred-envread
    $HomeDir = $env:USERPROFILE
  • install.ps1:539identity-write
    # AGENTS.md companion
  • install.ps1:544identity-write
    if (Test-Path (Join-Path $ScriptDir "AGENTS.md")) { return }
  • install.ps1:546identity-write
    Write-Info "[dry-run] Would generate companion AGENTS.md"
  • install.ps1:551identity-write
    $agentsMd = Join-Path $InstallDir "AGENTS.md"
  • install.ps1:566identity-write
    Write-Ok "Generated companion AGENTS.md"
  • scripts/evolve.py:236exec-spawn
    proc = subprocess.run(cmd, cwd=root)  # noqa: S603
  • scripts/review_staleness.py:43exec-spawn
    result = subprocess.run(
  • scripts/run_evals.py:313exec-spawn
    proc = subprocess.run(bound, shell=True, capture_output=True)  # noqa: S602
  • scripts/run_evals.py:313exec-shell-true
    proc = subprocess.run(bound, shell=True, capture_output=True)  # noqa: S602
  • scripts/run_evals.py:424exec-shell-true
    shlex.quote is POSIX-only: under Windows shell=True (cmd.exe) its single
  • scripts/run_evals.py:471exec-spawn
    proc = subprocess.run(  # noqa: S602
  • scripts/run_evals.py:472exec-shell-true
    bound, shell=True, cwd=str(skill_dir), capture_output=True,
  • scripts/run_evals.py:741cred-envread
    api_key = os.environ["ANTHROPIC_API_KEY"]
  • scripts/run_evals.py:760exec-spawn
    proc = subprocess.run(  # noqa: S603
  • scripts/run_evals.py:784cred-envread
    explicit = os.environ.get("EVAL_JUDGE_CMD", "").strip()
  • scripts/run_evals.py:804cred-envread
    if os.environ.get("ANTHROPIC_API_KEY"):
  • scripts/success_ledger.py:47cred-envread
    configured = os.environ.get("ASC_SUCCESS_LEDGER")
  • scripts/success_ledger.py:52cred-envread
    state_root = os.environ.get("XDG_STATE_HOME")
  • scripts/success_ledger.py:55cred-envread
    if os.name == "nt" and os.environ.get("LOCALAPPDATA"):
  • scripts/success_ledger.py:56cred-envread
    return Path(os.environ["LOCALAPPDATA"]) / "agent-skills-platform" / "success-events.jsonl"
  • skill.graph.json:17identity-write
    "path": "AGENTS.md",

这一栏是扫描器报的事实,不是结论。命中多不等于有毒(安全工具、规则库、示例脚本本来就会包含危险写法),命中少也不等于干净。它和你手上的凭据、文件、网络有什么关系,需要你自己看。

技能内容

/live-weather-briefing-skill

Create a current weather briefing for one city. Resolve the city through

Open-Meteo's geocoding API, then retrieve current conditions from its forecast API.

Always show the resolved location, observation time, and source URLs.

Run

python3 scripts/run_pipeline.py --city "Sao Paulo, BR" --output briefing.md

The command makes two live, read-only API requests and writes a Markdown briefing.

It never books travel, sends alerts, or changes external data.

Required behavior

  1. Use the first exact geocoding result and show its country and coordinates.
  2. Retrieve current temperature, apparent temperature, wind speed, weather code, and observation time.
  3. State that weather is advisory and include the source URLs.
  4. If a city cannot be resolved or an API response is incomplete, stop with an error; do not invent conditions.

Gotchas

  • Current conditions are model data, not a safety alert. Check local official warnings before decisions involving risk.
  • A city name can be ambiguous. Include a country or region in the request when the city is not unique.
  • Conditions change after retrieval; the briefing's observation time identifies the freshness of this result.

References

  • https://open-meteo.com/en/docs/geocoding-api
  • https://open-meteo.com/en/docs

想直接用这个技能?

本站把开放许可(MIT / Apache 等)的技能按仓库打包整理到网盘,点一下转存到你自己的网盘,不用一个个从 GitHub 拉。许可未声明的技能只给原始仓库链接,不打包。

它属于哪个仓库

星标★ 2,384
本站分层T2
该仓技能数2
原文件路径.agents/skills/live-weather-briefing-skill/SKILL.md

同一个仓库里的其他技能

看这个仓库的全部 2 个技能