Coding

Alibabacloud Cleversee Search

Try it

Alibaba Cloud cleversee CLI skill. web-search is a general-purpose, AI/agent-oriented internet search providing high-quality, low-latency neural knowledge re...

What it does

Alibaba Cloud cleversee CLI skill. web-search is a general-purpose, AI/agent-oriented internet search providing high-quality, low-latency neural knowledge retrieval and web content scraping across any topic and real-time information. Triggers: "cleversee", "web search", "AI 搜索", "联网搜索", "查资料", "实时信息".

The skill document

Scenario Description

cleversee is a standalone CLI tool for Alibaba Cloud users that provides:

  • Web Search (web-search): AI-powered internet search for agents — high-quality, low-latency retrieval of relevant results and page content across any topic and real-time information, with optional date-range and domain filtering. It creates no cloud resources and requires no cleanup.

Installation

Pre-check: cleversee CLI required

[MUST] Verify: cleversee --help — must succeed.

  • Install: npm install -g @cleversee/cli@1.0.5 --registry https://registry.npmjs.org

Environment Variables

cleversee does not require any environment variables. It reuses Aliyun CLI's credential profile system.

Authentication

Pre-check: Alibaba Cloud Credentials Required

Security Rules:

  • NEVER read, echo, or print AK/SK values (e.g., echo $ALIBABA_CLOUD_ACCESS_KEY_ID is FORBIDDEN)
  • NEVER ask the user to input AK/SK directly in the conversation or command line
  • ONLY use cleversee doctor to check credential status

Verification:

cleversee doctor

Check the output for profile validity.

If no valid profile exists: Configure credentials using cleversee auth set (see Authentication Configuration below).

Authentication Configuration

ModeInteractiveNon-Interactive Command
OAuthYes (browser required)Not supported
RamRoleArnNocleversee auth set --mode RamRoleArn --access-key-id --access-key-secret --arn

Interactive commands (cannot be used in non-interactive/automated contexts):

  • cleversee auth set (no parameters)
  • cleversee auth switch
  • cleversee auth --mode OAuth — OAuth login; skips the mode-selection prompt but still opens a browser, so it requires a present user

RAM Policy

This skill requires AliyunCleverSeeAISearchPlatformUserAccess permission. See references/ram-policies.md for the full permission list and policy document.

[MUST] Permission Failure Handling: When any command or API call fails due to permission errors at any point during execution, follow this process:

  1. Read references/ram-policies.md to get the full list of permissions required by this SKILL
  2. Use ram-permission-diagnose skill to guide the user through requesting the necessary permissions
  3. Pause and wait until the user confirms that the required permissions have been granted

Parameter Confirmation

IMPORTANT: Parameter Confirmation — Before executing any command or API call, ALL user-customizable parameters (e.g., search query, limit, search-type, region, date ranges, domain lists, profile name, etc.) MUST be confirmed with the user. Do NOT assume or use default values without explicit user approval.

Web Search Parameters

ParameterRequiredDefaultDescription
--query YesSearch keywords
--limit No10Result count, range 1-10
--search-type Nopropro (advanced, global sources) or lite (fast, low-cost)
--region Nomainland_chinamainland_china or global (pro mode only).
--start-time NoStart date YYYY-MM-DD
--end-time NoEnd date YYYY-MM-DD
--include-domain NoDomain whitelist, space-separated (pro only; forces --region global automatically)
--exclude-domain NoDomain blacklist, space-separated (pro only)
--tableNoOutput as a table view (default web-search view is a card list)

Global Parameters

Apply to web-search (and --profile also applies to auth). doctor accepts no options.

ParameterDefaultDescription
--profile defaultProfileAliyun CLI profile name
-o, --output tableOutput format (table for human-readable, json for program parsing)
-V, --versionPrint cleversee version

Core Workflow

Step 1: Verify environment     →  cleversee doctor
Step 2: Confirm search params  →  Ask user for query, limit, search-type, region, dates, domains
Step 3: Execute search         →  cleversee web-search --query "" [options] -o json
Step 4: Parse results          →  Check JSON output for at least one result

Example (parameters must be confirmed with user first):

# Step 1: Verify environment
cleversee doctor

# Step 2-3: Execute search with user-confirmed parameters
cleversee web-search \
  --query "What are the latest battery breakthroughs in electric vehicles?" \
  --limit 5 \
  --search-type pro \
  -o json

Workflow B: Credential Management

Step 1: Check current status   →  cleversee doctor
Step 2: Configure credentials  →  cleversee auth set --mode  [params]
Step 3: Verify                 →  cleversee doctor

Example:

# Step 1: Check current status
cleversee doctor

# Step 2: Configure RamRoleArn (parameters provided by user)
cleversee auth set --mode RamRoleArn \
  --access-key-id  \
  --access-key-secret  \
  --arn 

# Step 3: Verify
cleversee doctor

Success Verification

Quick checks:

  • cleversee doctor exits with code 0 and all checks pass
  • cleversee web-search exits with code 0 and returns at least one result

Cleanup

cleversee web-search is a read-only operation — it creates no cloud resources and requires no cleanup.

For credential cleanup, reconfigure or remove profiles via cleversee auth set or Aliyun CLI profile management.

Best Practices

  1. Run cleversee doctor first — always verify environment health before searching
  2. Use -o json — for programmatic result parsing
  3. Run cleversee doctor — to diagnose environment or credential issues when a command fails
  4. Use --search-type lite — for quick searches that don't need advanced filtering
  5. Use --include-domain / --exclude-domain — in pro mode to focus on authoritative sources
  6. Use --start-time / --end-time — to get recent content within a specific date range
  7. Use cleversee auth set for automation — never use interactive auth or auth switch in non-interactive contexts
  8. Use --profile — manage multiple account credentials
  9. Use --help — every command level supports --help for accurate parameter discovery
ReferenceDescription
references/ram-policies.mdRAM permission list and policy document
references/acceptance-criteria.mdAcceptance criteria with correct/incorrect patterns

Related skills

Real-time web search and URL reading with structured results backed by Aliyun IQS.

17 installs

Search and browse Alibaba Cloud agent skills via the AgentExplorer HTTP API, then install the ones that match your task.

45 installs

Run the AI Search Hub browser automation scripts for Yuanbao, LongCat, Doubao, Qwen, Gemini, and Grok. Use this skill when the user wants to ask one of those...

30 installs

Search Alibaba Cloud official help documentation (help.aliyun.com) with relevance-ranked search, and verify OpenAPI contracts (parameters, error codes, RAM permission points) against api.aliyun.com metadata. Use when the user asks how to use or configure an Alibaba Cloud product, looks up an error code or asks what an error message means, checks quota or usage limits, asks about billing rules, wants best practices or troubleshooting guides, confirms API parameter semantics, or wants to read a specific help document. Triggers: Alibaba Cloud documentation, help center, help.aliyun.com, product how-to guide, error code meaning, what does this error mean, how to fix this error, quota and limits, billing rules, RAM permission point, API reference, troubleshooting guide, best practice, read help document. Do not use this skill to execute changes on cloud resources, or to diagnose a specific product incident when a dedicated product diagnosis skill is installed and applicable.

WebScraping.AI (webscraping.ai). Use this skill for ANY WebScraping.AI request — searching and reading data. Whenever a task involves WebScraping.AI, use thi...

1 installs

支持 Bing/Baidu 通用搜索、知乎中文问答社区、小红书生活方式社区、GitHub代码仓库、 arXiv学术论文、199it互联网数据、东方财富财经资讯、智联招聘、前程无忧等多源搜索【按需持续增加】。 从多个平台搜索内容,实现统一搜索和内容提取,帮助AI Agent获取全网信息。 只要涉及获取网络信息——无论是搜索资讯、调研主题、查找资料、提取文章、多源对比—— 都应立即调用。即使用户未明确说"搜索",只要需要网络信息支撑,自动使用本skill。

1 installs1 stars