> ## Documentation Index
> Fetch the complete documentation index at: https://aisa.one/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# AIsa Tavily Search

> Search the web and extract readable page content through AIsa.

[View on GitHub ->](https://github.com/AIsa-team/agent-skills/tree/main/aisa-tavily)

**Web search and page extraction through AIsa.** Find current sources or turn public URLs into readable content with Tavily-backed endpoints.

## Install

First, install the AIsa CLI if you have not already:

```bash theme={null}
npm install -g @aisa-one/cli
```

Then install the skill:

```bash theme={null}
aisa skills install aisa-tavily
```

## What can agents do with it?

<CardGroup cols={2}>
  <Card title="Source discovery" icon="magnifying-glass">
    Find credible sources for a topic or question.
  </Card>

  <Card title="News lookup" icon="newspaper">
    Search current web results for recent context.
  </Card>

  <Card title="Page extraction" icon="scissors">
    Extract readable text from public URLs.
  </Card>

  <Card title="Research workflows" icon="file-lines">
    Gather sources before writing a brief or report.
  </Card>
</CardGroup>

## When to use

* When the user needs web search results for a topic, question, company, product, or event
* When the user wants source discovery before summarizing, comparing, or validating information
* When the user needs current-news lookup with recent-day filtering
* When the user provides a public URL and wants the page content extracted for downstream analysis

## When NOT to use

* Do not use this skill for sites that require login, browser interaction, cookies, or private account access
* Do not use this skill for posting, social engagement, media upload, or OAuth-based workflows; it performs search and public-URL extraction only
* Do not use this skill when there is no network access to `https://aisa.one` or `https://api.aisa.one`

## Quick Reference

| Task                       | Command                                                  |
| -------------------------- | -------------------------------------------------------- |
| Search the web             | `node scripts/search.mjs "query"`                        |
| Search with more results   | `node scripts/search.mjs "query" -n 10`                  |
| Run deeper research        | `node scripts/search.mjs "query" --deep`                 |
| Search news                | `node scripts/search.mjs "query" --topic news`           |
| Search recent news only    | `node scripts/search.mjs "query" --topic news --days 7`  |
| Extract content from a URL | `node scripts/extract.mjs "https://example.com/article"` |

## Capabilities

* Search the web through AIsa's Tavily-backed relay
* Return concise, relevant result sets for research and agent workflows
* Run deeper research with `--deep` for broader coverage
* Focus on news search with `--topic news`
* Limit news lookback windows with `--days <n>`
* Extract readable content from a public URL

## Search

```bash theme={null}
node scripts/search.mjs "query"
node scripts/search.mjs "query" -n 10
node scripts/search.mjs "query" --deep
node scripts/search.mjs "query" --topic news
```

## Options

* `-n <count>`: Number of results (default: 5, max: 20)
* `--deep`: Use advanced search for deeper research (slower, more comprehensive)
* `--topic <topic>`: Search topic - `general` (default) or `news`
* `--days <n>`: For news topic, limit to last n days

## Extract content from URL

```bash theme={null}
node scripts/extract.mjs "https://example.com/article"
```

## Setup

Requirements:

* `node`
* `AISA_API_KEY`
* Internet access with outbound requests to `https://aisa.one` and `https://api.aisa.one`

Auth, relay, and side-effect notes:

* This skill requires `AISA_API_KEY` from [https://console.aisa.one](https://console.aisa.one)
* Requests are sent through AIsa's remote relay at `https://aisa.one` and `https://api.aisa.one`
* This skill does not use OAuth
* This skill does not upload media or files
* This skill may send user search queries and public target URLs to the remote AIsa relay in order to return search results or extracted content

## Example Requests

* "Search for recent coverage of OpenAI enterprise pricing"
* "Find sources comparing vector databases for production RAG"
* "Look up this week's news about NVIDIA export controls"
* "Extract the main content from this article URL"

## Notes

* Needs `AISA_API_KEY` from [https://console.aisa.one](https://console.aisa.one)
* Powered by AIsa's unified API gateway (`https://aisa.one` / `https://api.aisa.one`)
* Use `--deep` for complex research questions
* Use `--topic news` for current events
* Search and extraction operate through remote relay requests, not local browser automation

## Get started

1. Sign up at [aisa.one](https://aisa.one) (new accounts start with \$2 free credit).
2. Generate an API key from the console.
3. Set your key and install the skill:
   ```bash theme={null}
   export AISA_API_KEY="your-key"
   npm install -g @aisa-one/cli
   aisa skills install aisa-tavily
   ```
4. Start a new agent session so the runtime loads the updated skill instructions.

## Related

<CardGroup cols={3}>
  <Card title="Tavily Search" icon="magnifying-glass" href="/agent-skills/tavily-search">
    Focused Tavily search with filters and depth controls.
  </Card>

  <Card title="Tavily Extract" icon="scissors" href="/agent-skills/tavily-extract">
    Extract clean readable content from URLs.
  </Card>

  <Card title="Multi-source Search" icon="magnifying-glass-plus" href="/agent-skills/search">
    Combine Tavily with scholar and Perplexity.
  </Card>
</CardGroup>