Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
SEARXNG_URLNoCustom SearXNG instance URL (fallback search engine)
BRAVE_API_KEYNoAPI key for Brave Search (enables search tool)

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": true
}
prompts
{
  "listChanged": true
}
resources
{
  "listChanged": true
}

Tools

Functions exposed to the LLM to take actions

NameDescription
fetchA

Fetch a public or authenticated URL as readable Markdown, with source, retrieval time, content hash and completeness. Specialized handlers support GitHub, Wikipedia, YouTube transcripts, PDFs, arXiv and tweets. Otherwise checks fresh shared content, then native Markdown/direct HTML, hosted rendering, and labeled archive fallbacks. Authenticated requests use direct sources only. Paginate long pages with startIndex and maxLength. noCache forces retrieval and disables archives unless explicitly allowed.

fetch_batchA

Fetch up to 10 URLs in parallel and return each as clean markdown. Same handler/fallback chain as the fetch tool, with a smaller per-URL length budget. Use after a search to pull several sources in one call.

researchA

Search the web and fetch the content of the top results in one call. Returns the full content of each result as markdown, ready to summarize or compare. Use this instead of separate search + fetch calls when researching a topic.

searchA

Search the web and return results. Uses Brave Search API if BRAVE_API_KEY is set, otherwise falls back to SearXNG and then DuckDuckGo. Supports domain filtering (site), freshness, and pagination (page).

extractA

Extract specific values from a web page as JSON instead of markdown prose. Provide CSS selectors to pull named fields (text or an attribute, first match or all), and/or set tables:true to convert every HTML table to arrays of row objects. Use this when you need particular data (prices, lists, specs, tabular data) rather than the whole page. Honors per-domain auth and proxies.

Prompts

Interactive templates invoked by user choice

NameDescription
research-topicSearch for a topic and fetch the top results. Provides a multi-source summary.
extract-articleFetch a URL and extract the key points from the content.

Resources

Contextual data attached and managed by the client

NameDescription
recent-fetchesURLs fetched and cached in this session, most recent first. Re-fetching any of these is instant.

TDQS

A4.1/5.0

Scored across 5 tools

Disambiguation4/5

Each tool has a distinct retrieval role: search (results only), fetch (single page as markdown), fetch_batch (parallel multi-URL), extract (CSS-targeted JSON), research (search+fetch combo). The only soft overlap is research versus chaining search+fetch, and fetch versus fetch_batch, but the descriptions explicitly steer usage and the boundaries are clear enough for an agent to select correctly.

Naming Consistency5/5

All names are lowercase single-verb or verb_noun forms (extract, fetch, fetch_batch, research, search), with consistent snake_case for the compound. No mixed conventions or casing drift.

Tool Count5/5

Five tools is well-scoped for a web retrieval server: one search, one fetch, a batch variant, a structured extractor, and a research convenience wrapper. No tool feels redundant or padded, and the surface stays small enough to reason about.

Completeness4/5

The core lifecycle of search → fetch → extract → batch → research is fully covered, including pagination, auth, caching, and fallbacks. Minor gaps exist around link/crawl discovery (e.g., enumerating links or sitemaps) and cache/status inspection, but agents can work around these with existing tools.

Maintenance

ActivityMaintained
ResponsivenessNo recent community issues