Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
AI Web Automation
v1.0.0Automate web tasks like form filling, data scraping, testing, monitoring, and scheduled jobs with multi-browser support and retry mechanisms.
⭐ 10· 14.7k·247 current·261 all-time
byZhangYang@arthasking123
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
high confidencePurpose & Capability
The SKILL.md claims multi-browser support (Selenium/Puppeteer), proxy pools, retry mechanisms, cron scheduling, email/notification integration and paid tiers. The code (main.py) only implements a simple requests.get-based scraper that saves a markdown report. package.json suggests a Node-style project but the runtime is a Python script. These requirements/features are disproportionate to the actual implementation and suggest overpromising or incomplete/misleading packaging.
Instruction Scope
Runtime instructions show CLI examples for many actions (scrape, cron, test, form-filling) and reference an 'openclaw run web-automation' interface with flags. The shipped main.py only accepts a positional 'scrape' action and a URL. The SKILL.md does not instruct reading unrelated local files or credentials, but it instructs actions that the code does not support — granting the agent broader authority in concept than the code exercises.
Install Mechanism
There is no install spec (instruction-only) which reduces install-time risk. However, package.json is present (Node-style) while the main executable is Python and imports 'requests' with no declared dependency; this mismatch is confusing and may lead to runtime surprises. No external download URLs or archives are used.
Credentials
The skill declares no required environment variables, no primary credential, and no config paths. The code does not reference environment secrets. This is proportionate to the actual minimal scraping functionality.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or system settings. It runs only when invoked and writes output to a local ./output directory. No elevated persistence is requested.
What to consider before installing
This package appears to overpromise in its README/SKILL.md compared with the actual code. Before installing or running: (1) Don't assume it contains Selenium, Puppeteer, proxy pools, or cron/email integrations — ask the author for the missing implementations. (2) Run it in a sandboxed environment if you need to test it; the script performs HTTP requests and writes files to ./output. (3) Ensure Python + 'requests' is installed; the package.json/Node hint is misleading. (4) If you need the advanced features, request an updated, auditable implementation (or refuse until the author provides code matching the advertised capabilities). (5) Avoid supplying credentials to this skill until its true behavior is clarified and verified.Like a lobster shell, security has layers — review code before you run it.
latestvk976dct954g0fracmf41z12qhs81g8ph
License
MIT-0
Free to use, modify, and redistribute. No attribution required.