Tavily
AI-powered web search, content extraction, crawling, and site mapping with Tavily API. Provides Search (basic/advanced web search with topic filtering), Extract (content extraction from URLs), Crawl (website crawling with depth control), and Map (site mapping to discover URLs). Supports API Key authentication. All nodes accept either a Client ID from Connect or direct credentials.
| Namespace | Robomotion.Tavily |
| Version | 1.1.0 |
| Category | Artificial Intelligence |
| Nodes | 7 |
Nodes
| Node | Description |
|---|---|
| Connect | Opens a connection to Tavily and returns a Client ID. |
| Disconnect | Closes a Tavily connection. |
| Search | Searches in Tavily. |
| Extract | Extracts in Tavily. |
| Crawl | Crawl — see the properties below. |
| Map | Map — see the properties below. |
| Toolkit | Exposes this package's operations to an AI Agent as a single tool. |
Connecting
This package has a Connect node. Run it once at the start of your flow; it returns a Client ID that the other nodes take as input. Pair it with Disconnect so the connection is released when the flow ends.
Credentials belong in a Vault rather than typed into the node.
AI Agent tools
6 of these nodes are exposed as tools an AI Agent can call directly:
Connect, Disconnect, Search, Extract, Crawl, Map.
Generated from the package spec that the Flow Designer itself loads, so the node list and every property below match the version above exactly.
It does not yet include worked examples or service-specific guidance — those are written by hand, and a plausible-sounding guess in a reference page is worse than an honest gap. If you build something useful with this package, the community is the place to share it.