← Back to Skills Marketplace
8584
Downloads
10
Stars
0
Active Installs
10
Versions
Install in OpenClaw
/install web-search-pro
Description
Agent-first web search and retrieval for live web search, news search, docs lookup, code lookup, company research, site crawl, site map, and structured evide...
Usage Guidance
Review this before installing in environments with access to internal HTTP services, cloud metadata, or valuable provider API keys. Use it only with trusted target URLs and trusted gateway configuration, avoid sensitive search text unless you accept provider transmission, and prefer disabling custom gateway overrides or running it in a network-restricted sandbox.
Capability Analysis
Type: OpenClaw Skill
Name: web-search-pro
Version: 2.1.4
The web-search-pro bundle is a legitimate and well-architected web search and research tool for AI agents. It implements a federated search system supporting multiple providers (Tavily, Exa, Brave, etc.) and includes a robust security boundary in `scripts/lib/url-safety.mjs` and `scripts/lib/web-fetch.mjs` to prevent SSRF attacks by blocking access to localhost, private IP ranges, and cloud metadata services. While `scripts/lib/http-client.mjs` utilizes `curl` via `execFile` for network requests, it does so without spawning a shell, and the overall logic is strictly aligned with the stated purpose of providing structured retrieval and research capabilities.
Capability Assessment
Credentials
Optional provider credentials are purpose-aligned, but custom Perplexity/OpenRouter gateway base URLs are accepted without clear scheme, destination, or private-network validation before sending bearer credentials and queries.
Install Mechanism
The install model is a bundled Node skill pack with no remote installer or curl-to-shell bootstrap in the inspected artifact.
Instruction Scope
The artifact advertises Safe Fetch blocking private and metadata targets, but the URL validation result is not pinned to the later fetch/curl connection, leaving a DNS-rebinding gap for extract, crawl, and map flows.
Persistence & Privilege
The skill writes local cache and provider health state under the configured cache directory; this is disclosed and scoped, with a cache clear command, and no system services or cross-session backdoors were found.
Purpose & Capability
The purpose is live web search, extraction, crawling, mapping, and research packs; outbound requests to DuckDuckGo, configured provider APIs, and user-supplied URLs are expected and mostly disclosed.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install web-search-pro - After installation, invoke the skill by name or use
/web-search-pro - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.1.4
Split docs by audience and language: English-only README, standalone README_zh, clearer agent-facing SKILL contract, and audited doc audience framing.
v2.1.3
Declared the Node install model, config.json read path, and .cache/web-search-pro state directory in ClawHub metadata and docs.
v2.1.2
Added clawdbot metadata, clarified the package as a code-backed Node runtime, and removed the Python helper transport from the no-key ddg baseline.
v2.1.1
Improve first-use onboarding with a real quick start and make federated-search gains visible through compact metrics like recovered results, corroborated results, and duplicate savings.
v2.1.0
Upgrade the search and routing substrate with search-router v2, bootstrap, cache telemetry, broader provider coverage, Perplexity gateway support, and head-to-head evaluation.
v2.0.1
Publish a ClawHub core profile with optional env disclosure and a slimmer registry package boundary.
v2.0.0
Web Search Pro 2.0.0 evolves the original 1.x multi-provider search skill into a retrieval stack with search, extract, render, crawl, map, controlled federation, runtime diagnostics, and structured research evidence packs.
v1.0.2
align metadata with multi-provider runtime and reduce false suspicious flagging
v1.0.1
metadata clarity and validation hardening
v1.0.0
- Initial release of Web Search Pro: a multi-engine, precision web search tool for AI agents.
- Supports Tavily, Exa, Serper, and SerpAPI with auto-selection based on query and available keys.
- Enables domain filtering, date ranges, deep/advanced search, news mode, and content extraction.
- Flexible parameter control: includes options like result count, language, and specific search engines.
- API-key based activation, with clear instructions for setup and usage.
- All features documented in English and Chinese.
Metadata
Frequently Asked Questions
What is Web Search Pro?
Agent-first web search and retrieval for live web search, news search, docs lookup, code lookup, company research, site crawl, site map, and structured evide... It is an AI Agent Skill for Claude Code / OpenClaw, with 8584 downloads so far.
How do I install Web Search Pro?
Run "/install web-search-pro" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Web Search Pro free?
Yes, Web Search Pro is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Web Search Pro support?
Web Search Pro is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Web Search Pro?
It is built and maintained by zjianru (@zjianru); the current version is v2.1.4.
More Skills