No user-submitted related servers found.
Scored across 29 tools
Tools are organized by source and action, but several tools (download_crossref, read_crossref, download_pubmed, etc.) return error messages instead of actual content, creating ambiguity. An agent may call these expecting results, leading to confusion. Additionally, the distinction between download_*, read_*, and get_* can be unclear for some sources.
Most tools follow a consistent verb_source or verb_source_paper pattern (e.g., search_arxiv, download_arxiv, read_arxiv_paper). Minor inconsistencies include 'get_crossref_paper_by_doi' vs. 'get_repec_paper' and some 'read_*' tools including 'paper' suffix while 'download_*' tools omit it. Overall, the pattern is predictable.
With 29 tools, the count is on the high side but largely justified by the number of sources and actions (search, download, read, metadata). However, several tools (e.g., download_crossref, read_pubmed) are essentially non-functional error messages, inflating the count. An effective tool count of about 20 would be more appropriate.
The server covers search, download, and full-text reading for several major sources, but notable gaps exist. Sources like PubMed and RePEc lack direct access, requiring users to rely on Sci-Hub or Semantic Scholar. The error-returning tools create dead ends, and there is no unified search across sources or citation metadata outside Semantic Scholar.