Equip your AI agents with real-time web search, image discovery, and LLM-friendly webpage extraction capabilities using the iFlow Search API.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install iflow-search
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install iflow-search using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is iFlow Search?
iFlow Search is an essential community skill designed to connect AI coding agents directly to the living web via the high-performance iFlow Search API. Instead of relying on static training data, agents can invoke simple bash scripts to retrieve live context, fetch relevant images, and download clean webpage text. By reducing noise and removing complex HTML boilerplate, it turns raw web data into token-friendly markdown.
As a crucial tool in the Openclaw Skills library, this skill helps developers build agents that perform automated market research, fact-check claims, and gather visual references. The interface operates completely via shell commands, making it incredibly straightforward to drop into any terminal-based agent workspace.
iFlow Search Use Cases
- Retrieval of current news or programming documentation beyond the AI model's knowledge cutoff.
- Image queries to find UI references, design patterns, or product images for coding tasks.
- Scrape known articles or websites to clean up boilerplate and supply clean content for document summarization.
- Multi-stage online workflows where the agent searches first, picks high-scoring links, and fetches each page sequentially.
How iFlow Search Works
- Initiation: The agent recognizes the need for live web search, image lookup, or webpage retrieval.
- Environment Validation: The shell script validates that
IFLOW_API_KEYis set and thatcurlandbashare available in the local PATH. - Query Execution: The agent calls
web_search.sh,image_search.sh, orweb_fetch.shwith custom query parameters. - API Fetch: The underlying curl commands trigger the iFlow API, which parses search results or processes target webpage structures.
- Token Delivery: Clean, structured results are printed to standard output for the AI model to ingest directly.
iFlow Search Setup
To get started with this skill, make sure you have bash and curl installed on your system. Then follow these configuration steps:
- Register and generate an API key on the iFlow Platform.
- Set the environment variable in your active shell profile (do not pass it as a command line argument directly):
echo 'export IFLOW_API_KEY="YOUR_IFLOW_API_KEY"' >> ~/.zshrc
source ~/.zshrc
- Verify your configuration by checking the variable before starting your agent:
echo "$IFLOW_API_KEY"
iFlow Search Data Schema & Taxonomy
This skill features three lightweight scripts that output clear structure formats:
Scripts and Capabilities
| Script | Purpose | Mandatory Arguments | Optional Arguments |
|---|---|---|---|
scripts/web_search.sh |
Queries web index | keywords |
num (default 10) |
scripts/image_search.sh |
Searches image index | keywords |
num (default 10) |
scripts/web_fetch.sh |
Retrieves & cleans a URL | url |
None |
Return Data Formats
Web Search: Returns an array or list of entries with key-value associations:
title: The header of the webpage.snippet: Brief text excerpt explaining the page content.link: The destination URL.
Image Search: Returns image records containing:
url: Direct path to the image asset.refUrl: Page URL hosting the image.title: Optional context title of the image.
Web Fetch: Returns a token-efficient JSON format with:
title: Page metadata title.content: Extracted article main text converted to clean markdown.url: Canonical source URL.fromCache: Boolean indicator showing if the system hit cached data.
iFlow Search Advanced Features
- Dynamic Two-Pass Search: Build automated scripts that perform search queries, analyze the returned URLs, and automatically run
web_fetch.shto compile highly contextual research documents. - High-Throughput Rate-Limiting Support: Features built-in handling for up to 1000 requests per minute (RPM) with predefined error codes (
USER_RATE_LIMIT) for resilient agent execution. - Boilerplate Filtering: Uses sophisticated server-side cleanup algorithms within iFlow Search API to bypass ads, navigation elements, and social media tracking codes, delivering pure text context.
- Error Diagnosis: Clear standard return codes like
PARAM_INVALID,INVALID_APIKEY, andSYSTEM_ERRORto help agents auto-resolve system anomalies or notify users correctly in custom Openclaw Skills workflows.
SKILL.md
Loading
Related Openclaw Skills
Comic Create
An AI-powered pre-production assistant that converts unstructured creative ideas into production-ready comic scripts, storyboards, and character designs.
Screenshot + Visual Analysis
A browser automation skill for capturing screenshots and performing visual reconnaissance to analyze pages for security indicators.
Audio Transcribe
A local, privacy-first audio transcription skill that converts speech to text, subtitles (SRT), or JSON using OpenAI's Whisper model.
AI-Native Founder Playbook
A bilingual, vendor-neutral operating playbook designed to convert AI-native startup ideas into stage-appropriate strategic experiments and automated agent workflows.
iOS SMSSDK Integration
An automated workflow skill designed to smoothly integrate MobTech SMSSDK into iOS projects with minimal disruption.
Trading Agents for Futures
A zero-API, rule-based six-dimensional analysis and decision engine for commodities futures markets, designed to power both AI agents and human traders.
Featured*
