SpecStream for CUGA
SpecStream is the live web data engine for teams and agents. IBM's open-source agent gets SpecStream from its tool registry. It searches the live web, reads whole pages and builds lists of records, with the price shown before any list runs.
What you get
- One entry in the registry YAML
- Add specstream under mcpServers with transport http and a bearer key read from the environment.
- Only the tools you list
- include keeps an unattended agent to search_web and read_pages.
- A price before a list runs
- Deep Search reads the request and quotes the most it can cost. Nothing paid starts until the maximum is approved.
Set it up
- Create an API key under Agents & API in your SpecStream workspace and export it as SPECSTREAM_API_KEY.
- Add https://api.specstream.ai/mcp under mcpServers with transport http and bearer auth.
- Start the registry with MCP_SERVERS_FILE pointing at your file, and ask.
The full setup guide has every step, with what to do if something doesn’t arrive.
Also in Agent frameworks & builders
The same live web data, in the other tools of this kind.
Live web data, in CUGA.
Ask in your own words for companies, places, people, filings or anything else the web lists.