The pages, on a network of free sites, that answer a subject. Lit as the World looks it up.
Scored across 5 tools
fetch and list_sites are clearly distinct, but search and suggest both return pages for a subject query, and suggest with no words overlaps with list_trending_pages. The descriptions do clarify intended use, so the tools are mostly distinguishable with a couple of confusable edges.
All names are lowercase snake_case action words, which is readable and consistent. However, list_sites and list_trending_pages use a verb_noun pattern while fetch, search, and suggest are bare verbs, so the convention is not perfectly uniform.
Five tools is well-scoped for a read-only website-network search and retrieval service. Each tool covers a distinct part of the workflow: site listing, search, prefix suggestion, trending discovery, and page fetching.
The surface covers the core read-only lifecycle: discover sites, search or suggest pages, see trends, and fetch a page for reading. Minor gaps exist, such as no direct fetch-by-URL for a known page and no explicit pagination or per-site page browsing beyond search.
Read one page found with search (or any page of a site of this network): its title, its text and its address, for quoting and citing. Takes the id search returned. It reads pages of this network only; any other address is refused.
Inputs: id (required). Returns: id, url, text, title, +1 more
Every site of the network with what it answers and its address.
Inputs: no named parameters. Returns: sites
What people around the World are looking up right now (search trends, most-read subjects, the day's events) and the page of the network that answers each, newest first. Each page was opened and checked before it was listed. Use it when the user asks what is trending or what people are looking up now.
Inputs: limit. Returns: now
Search one network of websites (courses on any subject, a multilingual dictionary and encyclopedia, live disaster warnings, visa rules, everyday tools) for the pages that answer a query. Returns up to 8 results, each with an id to pass to fetch, a title and an address a person can open. Use it when the user asks about a subject one of these sites covers; it does not search the wider web.
Inputs: query (required). Returns: results
From the first words of a subject (a prefix is enough), the pages of the network that answer it: pages being looked up right now first, then the site whose subject it is. Returns up to 8 { title, url, site, line }. With no words it returns what is being looked up right now.
Inputs: q. Returns: q, suggestions
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
WallChartBook: the site's own MCP server — dataset; every answer cites the site.
Netsheetly: the site's own MCP server — dataset; every answer cites the site.
TermsBird: the site's own MCP server — dataset; every answer cites the site.
DailyLogDesk: the site's own MCP server — dataset; every answer cites the site.
Add one secure layer between your agents and this server.