Tools
Kagi Search
Follow these steps to run the agent and tools in a cloud-hosted Meshagent Room.
1. Install the Meshagent CLI
2. Sign Up and Authenticate
Sign up and log in via the CLI:
Meshagent CLI Getting Started Guide
3. Set Your Environment Variables
KAGI_SUMMARIZER_ENGINE
- The summarization engine to use (e.g.,cecil
)KAGI_API_KEY
- Your Kagi API key (get one at Kagi)
You’ll pass these as --env
arguments during service startup.
4. Start the MCP Tool Service in a Room
- This starts a Meshagent Room with the MCP Server available as a tool.
- Note: Rooms automatically close if they go inactive.
5. Join the Room with a Chatbot Agent
- This starts a chatbot in the room with access to Kagi search and summarization tools.
- Multiple toolkits can be attached to the same agent or room.
- A link to the room will be provided in the output—use it to interact with the agent and tools.
Environment Variables
Variable | Description |
---|---|
KAGI_SUMMARIZER_ENGINE | Summarization engine (e.g., cecil ) |
KAGI_API_KEY | Your Kagi API key |
Tools Available
Tools Provided
Tool Name | Short Description |
---|---|
kagi_search_fetch | Fetch web results based on one or more queries using the Kagi Search API. |
kagi_summarizer | Summarize content from a URL using the Kagi Summarizer API. |
Tools Details
Tool: kagi_search_fetch
Fetch web results based on one or more queries using the Kagi Search API. Use for general search and when the user explicitly tells you to “fetch” results/information. Results are from all queries given. They are numbered continuously, so that a user may refer to a result by a specific number.
- Parameters:
queries
(array
): One or more concise, keyword-focused search queries. Include essential context within each query for standalone use.
Tool: kagi_summarizer
Summarize content from a URL using the Kagi Summarizer API. The Summarizer can handle any document type (text webpage, video, audio, etc.).
- Parameters:
url
(string
): A URL to a document to summarize.summary_type
(string
, optional): Type of summary to produce. Options are'summary'
(paragraph prose) and'takeaway'
(bulleted list).target_language
(string
, optional): Desired output language (e.g.,'EN'
). If not provided, the original document language is used.
Meshagent and Kagi MCP Tool Resources
- Meshagent Website
- Meshagent Room API: Getting Started
- Meshagent Room API Overview
- Meshagent MCP Kagi Search Docker Image
Run your Kagi search & summarization tools interactively, collaboratively, and scalably—with Meshagent Rooms!