← Back to Skills Marketplace
565
Downloads
0
Stars
1
Active Installs
2
Versions
Install in OpenClaw
/install websearch-free-skill
Description
Perform web searches using various search engines and APIs.
Usage Guidance
This package's documentation describes Node scripts and npm dependencies (axios, mcporter) and instructs you to run 'npm install' and 'node search.js', but the registry entry only contains SKILL.md and README.md — the actual code and package.json are missing. Before installing or running anything: (1) ask the publisher for the missing source files or a link to the project repo/homepage and verify package.json and code; (2) inspect package.json and node modules for unfamiliar or high-risk packages (mcporter is uncommon); (3) avoid running npm install or executing downloaded scripts until you've reviewed them; (4) if you must test, do so in a strict sandbox or ephemeral environment with no sensitive data or credentials available. The mismatch between claimed behavior and the provided files is the reason this is flagged as suspicious.
Capability Analysis
Type: OpenClaw Skill
Name: websearch-free-skill
Version: 0.1.1
The skill provides legitimate web search and content extraction capabilities using the Jina Reader API (r.jina.ai) and Exa search engine via the mcporter framework. The implementation in search.js and exasearch.js is transparent, lacks obfuscation, and contains no indicators of data exfiltration, unauthorized execution, or malicious prompt injection.
Capability Assessment
Purpose & Capability
Name/description match the claimed capability (web search and page reading) and the required binaries (node, npm) are consistent with a Node.js implementation — that part is coherent. However, the skill documentation lists runtime scripts (search.js, exasearch.js), a package.json/dependencies (axios, mcporter) and an npm install step, but the asset manifest contains only README.md and SKILL.md (no code files, no package.json). Requesting Node/npm without shipping the code or a package.json is disproportionate and incoherent.
Instruction Scope
SKILL.md instructs the agent/operator to run node ./search.js and node ./exasearch.js to fetch arbitrary web URLs and query a search provider (Jina Reader API and Exa via mcporter). Fetching external URLs is expected for a websearch skill, but the instructions also direct automatic dependency installation via npm and call out the third-party 'mcporter' integration. Because the actual script files are not present in the package, the instructions are incomplete and grant broad discretion (install and run arbitrary npm packages) that is outside the provided artifact.
Install Mechanism
No install spec is provided in the registry (instruction-only), which is the lowest-risk pattern — but SKILL.md explicitly tells users to run 'npm install' and lists dependencies. That mismatch is concerning: the skill asks to install packages at runtime even though no package.json or code is included. Installing arbitrary npm dependencies (including 'mcporter') without a vetted source increases risk.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. For a websearch/URL-reading skill this is proportionate. There is no evidence the skill requests unrelated secrets.
Persistence & Privilege
Flags show always: false and normal autonomous invocation settings. The skill does not request permanent presence or elevated agent-wide privileges.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install websearch-free-skill - After installation, invoke the skill by name or use
/websearch-free-skill - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.1
**Major breaking change: All core source files removed; SKILL.md now includes only documentation and sample code.**
- Removed main logic files: ReadResult.js, search.js, exasearch.js, package.json, and config/mcporter.json.
- SKILL.md updated to embed sample code and keep documentation only; no executable scripts included.
- Usage instructions and error handling sections remain as reference.
- Skill now serves as a documentation/sample reference rather than a working implementation.
v0.1.0
- Initial release of websearch skill for performing web searches without needing API keys for basic usage.
- Supports extracting content from any web page URL and performing keyword-based searches with customizable result limits.
- Automatic dependency installation via npm; requires Node.js v22+.
- Provides detailed error handling for missing parameters, invalid values, network issues, and HTTP errors.
- Utilizes Jina Reader API for content extraction and Exa API (via mcporter) for search functionality.
Metadata
Frequently Asked Questions
What is Websearch?
Perform web searches using various search engines and APIs. It is an AI Agent Skill for Claude Code / OpenClaw, with 565 downloads so far.
How do I install Websearch?
Run "/install websearch-free-skill" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Websearch free?
Yes, Websearch is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Websearch support?
Websearch is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Websearch?
It is built and maintained by Rory (@roryyu); the current version is v0.1.1.
More Skills