How do I install the Multi Search Engine skill?

Run npx skillmds@latest add dvcrn/multi-search-engine-2 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Multi Search Engine safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: CAUTION, Skill Scanner: PASS. Capability flags: executes scripts. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Multi Search Engine?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Multi Search Engine free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Multi Search Engine?

dvcrn (@dvcrn) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Multi Search Engine by dvcrn — Agent Skill

Name: Multi Search Engine
Rating: 4.2 (5 reviews)
Author: dvcrn

File contents

Multi Search Engine v2.1.0

A search orchestration skill for OpenClaw that helps the agent choose engines, build safe search URLs, compare results across engines, and explain why a given engine is appropriate.

This version turns the skill from a static engine list into a reusable toolkit with:

engine aliases and region-aware selection
advanced operator composition (site:, filetype:, exact phrase, exclusion, OR groups)
time scopes where the engine supports them
compare mode for multi-engine research
privacy-first presets
command-line helper script that outputs json, markdown, or plain text
explicit safety guidance for direct web access

When to use this skill

Use this skill when the user wants to:

search the same query on multiple engines
bias toward Chinese vs global engines
use privacy-first engines
build advanced search operators reliably
compare engine coverage before opening pages
prepare direct web_fetch or browser URLs without needing an API key

Recommended workflow

Normalize the user's research intent:
- broad web search
- China-focused search
- privacy-first search
- docs / code / paper / file search
- calculation / fact query
Pick engines deliberately:
- China / local web: Baidu, Sogou, 360, Toutiao, WeChat
- Global general: Google, Bing INT, Brave, Yahoo
- Privacy-first: DuckDuckGo, Startpage, Brave, Qwant
- Knowledge / computation: WolframAlpha
- Community / niche finance: Jisilu
Build the query using operators only when they help:
- site: for source restriction
- filetype: for PDFs / docs
- quotes for exact phrase
- -term for exclusion
- OR for alternatives
Prefer compare mode for research-sensitive tasks:
- one privacy engine
- one mainstream engine
- one region-specific engine if relevant
Open only the most promising result pages after inspecting generated URLs.

Direct script usage

python3 scripts/build_search_urls.py --query "openclaw skills" --engine google
python3 scripts/build_search_urls.py --query "量化投资" --preset cn-research --time week --format markdown
python3 scripts/build_search_urls.py --query "react hooks" --site github.com --exact "useEffect" --compare google,ddg,brave
python3 scripts/build_search_urls.py --query "100 USD to CNY" --engine wolframalpha --format json

Typical OpenClaw usage

User asks for recent AI papers as PDFs
→ build a Google / Brave compare query
→ add filetype:pdf and time scope
→ inspect URLs before opening pages

Security and privacy

This skill does not require API keys.
It does not execute remote code.
It only generates deterministic search URLs from local templates.
Network access happens only when the agent explicitly opens the generated URLs.
Search engines may log requests; use privacy presets when that matters.

Files

SKILL.md — primary instructions and metadata
README.md — install, scenarios, examples, FAQ, risks
scripts/build_search_urls.py — URL builder and compare helper
resources/engine-catalog.json — engine definitions, aliases, capabilities
resources/search-operator-cheatsheet.md — operator reference
examples/example-prompt.md — prompt patterns for agents
tests/smoke-test.md — manual smoke coverage
SELF_CHECK.md — self-audit and release checklist

---name: multi-search-engine-2description: Build auditable search URLs across Chinese and global engines with region/language filters, advanced operators, time scopes, privacy-first options, compare mode, and no API keys.--- # Multi Search Engine v2.1.0 A search orchestration skill for OpenClaw that helps the agent **choose engines, build safe search URLs, compare results across engines, and explain why a given engine is appropriate**. This version turns the skill from a static engine list into a **reusable toolkit** with: - engine aliases and region-aware selection - advanced operator composition (`site:`, `filetype:`, exact phrase, exclusion, OR groups) - time scopes where the engine supports them - compare mode for multi-engine research - privacy-first presets - command-line helper script that outputs `json`, `markdown`, or plain `text` - explicit safety guidance for direct web access ## When to use this skill Use this skill when the user wants to: 1. search the same query on multiple engines 2. bias toward Chinese vs global engines 3. use privacy-first engines 4. build advanced search operators reliably 5. compare engine coverage before opening pages 6. prepare direct `web_fetch` or browser URLs without needing an API key ## Recommended workflow 1. Normalize the user's research intent: - broad web search - China-focused search - privacy-first search - docs / code / paper / file search - calculation / fact query 2. Pick engines deliberately: - **China / local web**: Baidu, Sogou, 360, Toutiao, WeChat - **Global general**: Google, Bing INT, Brave, Yahoo - **Privacy-first**: DuckDuckGo, Startpage, Brave, Qwant - **Knowledge / computation**: WolframAlpha - **Community / niche finance**: Jisilu 3. Build the query using operators only when they help: - `site:` for source restriction - `filetype:` for PDFs / docs - quotes for exact phrase - `-term` for exclusion - `OR` for alternatives 4. Prefer compare mode for research-sensitive tasks: - one privacy engine - one mainstream engine - one region-specific engine if relevant 5. Open only the most promising result pages after inspecting generated URLs. ## Direct script usage ```bash python3 scripts/build_search_urls.py --query "openclaw skills" --engine google python3 scripts/build_search_urls.py --query "量化投资" --preset cn-research --time week --format markdown python3 scripts/build_search_urls.py --query "react hooks" --site github.com --exact "useEffect" --compare google,ddg,brave python3 scripts/build_search_urls.py --query "100 USD to CNY" --engine wolframalpha --format json ``` ## Typical OpenClaw usage ```text User asks for recent AI papers as PDFs → build a Google / Brave compare query → add filetype:pdf and time scope → inspect URLs before opening pages ``` ## Security and privacy - This skill does **not** require API keys. - It does **not** execute remote code. - It only generates deterministic search URLs from local templates. - Network access happens only when the agent explicitly opens the generated URLs. - Search engines may log requests; use privacy presets when that matters. ## Files - `SKILL.md` — primary instructions and metadata - `README.md` — install, scenarios, examples, FAQ, risks - `scripts/build_search_urls.py` — URL builder and compare helper - `resources/engine-catalog.json` — engine definitions, aliases, capabilities - `resources/search-operator-cheatsheet.md` — operator reference - `examples/example-prompt.md` — prompt patterns for agents - `tests/smoke-test.md` — manual smoke coverage - `SELF_CHECK.md` — self-audit and release checklist

Frequently asked questions

Multi Search Engine