Terminal
Terminal/Skill
Searches the web and reads page content through Alibaba Cloud IQS, returning results with source links for current facts and research.
By Alibaba Cloud
Works with
npx skills add aliyun/alibabacloud-aiops-skills --skill alibabacloud-iqs-searchreadpagesearchFollow the best practices to determine parameter values. Use default values when uncertain:
Follow the best practices to determine parameter values. Use default values when uncertain:
You MUST execute the scripts via bash command (e.g., node scripts/search.mjs ... or node scripts/readpage.mjs ...). Do NOT use your built-in web_search, WebFetch, or any other internal tools as substitutes. If the script fails, retry or report the error — do NOT fall back to built-in tools.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
--query | string | Yes | - | Search query (1-500 chars) |
--engineType | string | No | LiteAdvanced | Search engine type |
--timeRange | string | No | NoLimit | Time range filter |
--contents | string | No | - | Type of return content |
--numResults | int | No | 10 | Number of search results (1-10) |
1. Query Optimization (--query)
2. Engine Selection (--engineType)
The four engines differ significantly in latency, recall depth, content length, and cost. Pick the cheapest engine that meets the task — do NOT default to Deep; it is ~10× slower and 50× more expensive than the standard engine.
| Engine | Avg RT | Result count | snippet | mainText | Advanced filters | Multilingual | Cost ratio (vs Standard) | Use case |
|---|---|---|---|---|---|---|---|---|
Generic | ~950ms | ~10 | ~150 chars | ≤3000 chars | ✗ | Medium | 1× | General search, news/realtime, scene queries like weather (supports city/ip) |
LiteAdvanced | ~500ms | 1-50 | ~500 chars | ≤3000 chars | ✓ | Good | Lite-tier 1× | Default recommendation: low-latency semantic search; snippets are already rich enough |
Deep | ~6s | 1-50 | ≤500 chars | ≤50000 chars | ✓ | Good (CN/EN) | Vertical 50× | Complex multi-step reasoning, research reports, offline/Agent tasks needing deep browsing |
Decision rules:
LiteAdvanced: low latency + semantic search + snippet covers most Agent needs without needing extra mainText fetches.Generic when: the query is short and clearly informational (news, weather, simple facts), or when minimizing cost matters; also the only engine honoring city / ip for scene results (weather etc.).Deep ONLY when: the question is multi-hop / complex reasoning (FRAMES/BrowseComp-style), OR you need very long mainText (≤50000 characters, ~16× longer than other engines) for downstream LLM reasoning.
Deep for: real-time chat, simple lookups, high-QPS scenarios — latency (~6s) and cost (50×) are prohibitive.3. Time Range Selection (--timeRange)
NoLimit: Default when uncertain - engine optimizes based on query relevanceOneDay: Today onlyOneWeek: Last 7 daysOneMonth: Last 30 daysOneYear: Last 365 days4. Content Return (--contents)
mainText: Return full main text content - Use when detailed information is needed, such as technical documentation, research reports, or in-depth articlessummary: Return concise summary only - Use when a quick overview is sufficient, or when the page content is too large and token reduction is needed5. Result Count (--numResults)
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
--url | string | Yes | - | Target page URL |
--format | string | No | markdown | Return format |
--timeout | number | No | 60000 | Total timeout in milliseconds |
--pageTimeout | number | No | 15000 | Page load timeout in milliseconds |
--stealth | number | No | 0 | Enable stealth mode (0 or 1) |
--extractArticle | boolean | No | false | Extract main article content only |
1. Format Selection (--format)
markdown: Best for articles, preserves structure (default)text: Best for data extractionhtml: When structure analysis needed2. Article Extraction (--extractArticle)
3. Handling Failures (--timeout, --stealth)
--timeout value--stealth 1node scripts/search.mjs --query "量子计算原理" --engineType LiteAdvanced
node scripts/search.mjs --query "最新金融政策" --engineType Generic --timeRange OneWeek
node scripts/search.mjs --query "www.aliyun.com" --engineType LiteAdvanced --numResults 3
node scripts/search.mjs --query "AI 法案" --engineType LiteAdvanced --contents mainText
node scripts/search.mjs --query "人工智能行业年度报告" --engineType LiteAdvanced --contents summary
# Returns up to 50000-char mainText per result; latency ~6s. Default timeout auto-bumps to 60s for Deep.
node scripts/search.mjs --query "对比 GPT-5 与 Claude Opus 4.7 的代码能力差异" --engineType Deep --numResults 5 --contents mainText
node scripts/readpage.mjs --url "https://example.com/article" --format markdown --extractArticle true
node scripts/readpage.mjs --url "https://example.com/article" --format text --timeout 60000
node scripts/readpage.mjs --url "https://example.com/article" --format markdown --stealth 1 --extractArticle true
After executing any search.mjs or readpage.mjs command:
ls -la <filepath> and head -20 <filepath> to confirm the file exists and contains valid data.If the script returns an error about missing API key:
STOP the current task immediately. Do NOT fall back to built-in tools (WebFetch, web_search, curl, etc.) as substitutes.
Report the error to the user and ask the user to configure the API key:
Retry the task with following instruction: Method 1: Environment Variable
export ALIYUN_IQS_API_KEY="your-api-key"
Method 2: Configuration File
Create or edit ~/.alibabacloud/iqs/env:
ALIYUN_IQS_API_KEY=your-api-key