Surf Query

查詢台灣全島衝浪浪點資訊，含即時潮汐、風況、颱風動態、日出日落，可附帶查詢附近停車場。支援 Telegram、LINE、iMessage。

This skill appears to implement the surf/tide/wind features it advertises, but take these precautions before installing: - Provide a CWA_API_KEY only if you need live tides/wind/typhoon data; the SKILL.md explains how to add it. Note the registry metadata omitted this requirement — that mismatch is just a documentation issue, not a showstopper, but be aware. - Inspect the code before enabling automatic updates: update.sh will curl raw files from a GitHub repo and overwrite the local skill. If you rely on this, consider locking updates to a specific commit or reviewing fetched files before replacing local code. - The code contains calls that disable TLS verification for CWA endpoints (verify=False + disable_warnings). This reduces security and could enable MITM attacks. If possible, ask the maintainer for a fix, or only run live queries on a trusted network. - If you plan to enable cron-based push notifications, confirm where messages will be sent and which channel bindings are configured in OpenClaw (Telegram/LINE/iMessage). These are expected features but can send notifications to external accounts. If you are comfortable reviewing the script yourself and either removing or accepting the verify=False workaround and the auto-update behavior, the skill is coherent with its purpose. If you cannot review code, consider running it in a constrained environment or asking the maintainer for a signed release or pinned update mechanism.

Type: OpenClaw Skill Name: surf-query Version: 1.2.0 The skill is classified as suspicious primarily due to two significant security vulnerabilities. First, the `update.sh` script performs an auto-update by directly downloading and overwriting `surf_query.py`, `skill.yml`, and `taiwan_surf_spots.json` from a GitHub repository (`https://github.com/Harperbot/openclaw-surf-query/raw/main/`). This introduces a supply chain risk, as a compromise of the GitHub repository could lead to the automatic execution of malicious code on the user's system (seen in `update.sh`, `SKILL.md`, `README.md`). Second, the `surf_query.py` script and example code in `README.md` disable SSL/TLS certificate verification (`verify=False`) when making requests to the CWA API, making these communications vulnerable to Man-in-the-Middle attacks. While the skill's stated purpose is benign and there's no direct evidence of intentional malice (e.g., data exfiltration to unauthorized endpoints, explicit prompt injection against the agent), these vulnerabilities present significant security risks.