← 返回 Skills 市场
vijay-murugan

Houston Transtar Watch

作者 vijay-murugan · GitHub ↗ · v1.0.0
cross-platform ⚠ suspicious
377
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install houston-transtar-watch
功能描述
Poll Houston TranStar incidents RSS every 10 minutes and WhatsApp me when there are changes.
使用说明 (SKILL.md)

Houston TranStar Watcher

Overview

This skill is a specialized traffic monitor for the Greater Houston area. It interfaces with the Houston TranStar Real-Time Incident RSS feed to detect new accidents, stalled vehicles, and road closures.

Technical Runbook

The following logic is executed every 10 minutes via the internal cron scheduler:

  1. Script Execution: The skill navigates to the local directory and executes the diff-check logic:
    python3 transtar_diff.py
    
  2. Output Parsing: - If the script returns the string NO_CHANGES, the process terminates immediately to save bandwidth and prevent notification fatigue.

If any other text is returned, it is treated as a "Delta Report" of new incidents. 3. Delivery Logic When a "Delta Report" is generated, the content is formatted for mobile viewing and pushed via the WhatsApp Gateway to the configured recipient.

  1. Use Cases Automated commute monitoring for Houston residents.

Alerting logistics teams to major freeway closures on I-10, I-45, and US-59.3.

安全使用建议
This skill contains working RSS-fetching and diff logic, but it is incomplete and misconfigured for notifications. Before installing or enabling it: 1) Do not provide any WhatsApp/API credentials until you see explicit, secure code that uses them — the package currently has no delivery implementation. 2) Fix the mismatches: SKILL.md calls transtar_diff.py but the repo contains transtar.py; main.py points at scripts/your_script.py — confirm which script is the real entrypoint and update the loader. 3) Ensure the runtime has the 'requests' library or add a dependency/install step. 4) Review where state is written (~/.openclaw/skills/houston-transtar-watch/state/last_incidents.json) and decide if that location and persistence are acceptable. 5) If you want WhatsApp delivery, prefer using a well-known gateway (e.g., Twilio) and ensure credentials are provided via secure environment variables (and documented in the skill) rather than hardcoded. If the maintainer provides an updated package with clear delivery code and documented, minimal env vars, the assessment could be upgraded to benign.
功能分析
Type: OpenClaw Skill Name: houston-transtar-watch Version: 1.0.0 The skill is suspicious due to a critical misconfiguration in `main.py`. It attempts to execute a non-existent script named `your_script.py` via `subprocess.run` using a relative path (`../../scripts/your_script.py`). This path resolution is unusual and could lead to arbitrary code execution if an attacker could place a malicious `your_script.py` in the resolved directory. Additionally, the `SKILL.md` instructs the agent to run `transtar_diff.py`, which also does not exist, while the actual functional code is in `transtar.py` (present twice, once in `scripts/` and once at the root). This combination of non-existent script execution and path ambiguity represents a significant vulnerability, even though the `transtar.py` script itself appears benign.
能力评估
Purpose & Capability
The skill claims to poll the Houston TranStar RSS and send WhatsApp notifications. The included Python scripts (transtar.py and scripts/transtar.py) do fetch the RSS and write state to ~/.openclaw/... which is consistent with the monitoring purpose. However, there is no code that implements WhatsApp delivery or any declared credentials for a WhatsApp gateway (e.g., Twilio, Meta) and SKILL.md mentions pushing via a WhatsApp Gateway but provides no configuration. The entrypoint in main.py points to a placeholder path (scripts/your_script.py) and SKILL.md instructs running transtar_diff.py — neither of those files exist, so the pieces required to perform the stated end-to-end task are missing or mismatched.
Instruction Scope
SKILL.md instructs running a diff-check script every 10 minutes and sending a 'Delta Report' via WhatsApp. The scripts present produce the expected 'NO_CHANGES' or a summary and save state locally, which is within scope. But the SKILL.md references filenames (transtar_diff.py) that are not in the bundle, and it says to push via a WhatsApp gateway without specifying how or what credentials to use. The instructions are therefore vague and incomplete, granting broad discretion (e.g., how to configure delivery) and lacking concrete, safe steps for notification delivery.
Install Mechanism
There is no install spec (instruction-only style), so nothing external will be downloaded or written by an installer. The scripts import the third-party 'requests' library but do not declare dependencies; this may cause runtime failures if the environment lacks requests. The scripts write a state file under the user's home directory (~/.openclaw/.../state/last_incidents.json), which is expected for this functionality but is persistent disk I/O the user should be aware of.
Credentials
The skill declares no required environment variables or primary credential, yet the SKILL.md promises WhatsApp delivery (which normally requires API keys, tokens, or webhook configuration). No credentials are requested or documented, and there is no code to read or send WhatsApp messages. This mismatch means the skill as packaged cannot complete its claimed delivery step and may require the user to supply sensitive credentials later or modify code to add them — an opportunity for misconfiguration or accidental credential exposure.
Persistence & Privilege
The skill does not request always:true and is user-invocable; it stores its own state under ~/.openclaw/... which is reasonable for a polling notifier. It does not modify other skills or system-wide settings. Autonomous invocation is allowed (disable-model-invocation is false) which is the platform default and appropriate for scheduled polling.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install houston-transtar-watch
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /houston-transtar-watch 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
- Initial release of houston-transtar-watch skill. - Polls Houston TranStar incidents RSS feed every 10 minutes. - Detects and reports new accidents, stalls, and road closures. - Notifies user via WhatsApp with a summary of incidents when changes occur. - Designed for automated traffic monitoring and alerting in the Greater Houston area.
元数据
Slug houston-transtar-watch
版本 1.0.0
许可证
累计安装 1
当前安装数 1
历史版本数 1
常见问题

Houston Transtar Watch 是什么?

Poll Houston TranStar incidents RSS every 10 minutes and WhatsApp me when there are changes. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 377 次。

如何安装 Houston Transtar Watch?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install houston-transtar-watch」即可一键安装,无需额外配置。

Houston Transtar Watch 是免费的吗?

是的,Houston Transtar Watch 完全免费(开源免费),可自由下载、安装和使用。

Houston Transtar Watch 支持哪些平台?

Houston Transtar Watch 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Houston Transtar Watch?

由 vijay-murugan(@vijay-murugan)开发并维护,当前版本 v1.0.0。

💬 留言讨论