← Back to Skills Marketplace
1194
Downloads
0
Stars
8
Active Installs
2
Versions
Install in OpenClaw
/install selenium-browser-skill
Description
Teaches the agent how to perform advanced web automation using Python, Selenium WebDriver, and ChromeDriver.
Usage Guidance
This skill appears internally consistent and focused on Selenium-based browser automation, but take these precautions before installing or using it: 1) Review any Python script the agent generates before granting permission to run it — the SKILL.md correctly requires explicit approval. 2) Do not provide site credentials, API keys, or other secrets to the agent unless you trust the execution environment; automated scripts can access any page content and could capture sensitive data. 3) Ensure Chrome and ChromeDriver are installed from trusted sources (package manager or official releases). 4) If your platform allows restricting skill capabilities, consider limiting exec/write or requiring manual approval for execution to enforce the SKILL.md rule. 5) Run generated scripts in an isolated environment (VM/container) if they will browse untrusted sites. If you want a deeper check, provide a sample generated script and the platform policies for exec/write so I can verify the script matches the stated safeguards.
Capability Analysis
Type: OpenClaw Skill
Name: selenium-browser-skill
Version: 0.1.1
The skill is classified as suspicious due to its requirement for high-risk permissions (`exec`, `write`) and its ability to generate and execute arbitrary Python code, including the use of Selenium's `execute_script` for arbitrary JavaScript injection into web pages. While these capabilities are necessary for its stated purpose of web automation, they present a significant attack surface. However, the `SKILL.md` explicitly instructs the AI agent to 'Never run the script automatically' and to 'ask the user for explicit permission to run it' before using the `exec` tool, which mitigates the risk of autonomous malicious actions by the agent. There is no direct evidence of intentional malicious behavior (e.g., data exfiltration, persistence) designed into the skill bundle itself, making it a powerful tool with potential for misuse rather than inherently malicious code.
Capability Assessment
Purpose & Capability
The name and description match the instructions: SKILL.md teaches Python + Selenium usage and lists the expected system/python dependencies (python3, chromedriver, google-chrome, selenium). The frontmatter permission requirements (exec, write) align with a skill that writes scripts and may run them after user approval.
Instruction Scope
The runtime instructions stay on-topic (creating scripts, navigation, screenshots, JS execution, closing the browser). The SKILL.md explicitly instructs the agent NOT to run scripts without user approval and to only use exec after explicit consent — a deliberate safety measure. One notable capability: examples include driver.execute_script (JavaScript injection) which is a normal automation tool but can be used to read/modify page DOM and could be misused to collect or exfiltrate sensitive page data if a generated script sends that data elsewhere; the skill itself does not instruct exfiltration.
Install Mechanism
This is an instruction-only skill; the only install guidance inside the frontmatter is a pip install selenium line. There are no downloads from arbitrary URLs or archive extraction. The skill assumes the user or environment supplies Chrome/ChromeDriver; it does not attempt to fetch them itself.
Credentials
The skill requests no environment variables or external credentials. It does request exec and write permissions (to create and optionally run scripts), which are proportionate for a tool that generates and executes automation scripts.
Persistence & Privilege
always is false and the skill is user-invocable. The skill's required capabilities (exec, write) allow writing and running scripts but it does not request elevated persistent presence or modify other skills or system-wide configuration. Autonomous invocation is possible per platform defaults, but there are no additional privileges requested that would materially increase risk.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install selenium-browser-skill - After installation, invoke the skill by name or use
/selenium-browser-skill - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.1
- Added explicit security and execution rules: scripts must not be run automatically; user approval is now required before executing automation code.
- Documented system and Python dependencies (python3, chromedriver, google-chrome, and selenium) in the skill metadata.
- Included installation instructions for required Python packages.
- No functional code changes were made to the Selenium usage examples or guidance.
v0.1.0
- Initial release of the selenium-automation skill.
- Provides step-by-step instructions and code snippets for advanced web automation using Python and Selenium WebDriver with ChromeDriver.
- Covers essential operations: browser setup (headless mode), navigation, element interaction, JavaScript execution, taking screenshots, and safe browser closure.
- Designed for seamless automation, scraping, and browser scripting tasks.
Metadata
Frequently Asked Questions
What is Selenium Automation?
Teaches the agent how to perform advanced web automation using Python, Selenium WebDriver, and ChromeDriver. It is an AI Agent Skill for Claude Code / OpenClaw, with 1194 downloads so far.
How do I install Selenium Automation?
Run "/install selenium-browser-skill" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Selenium Automation free?
Yes, Selenium Automation is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Selenium Automation support?
Selenium Automation is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Selenium Automation?
It is built and maintained by gg-erick (@gg-erick); the current version is v0.1.1.
More Skills