MCP server for Elsevier Developer APIs (Scopus, ScienceDirect, and more)
14 tools with complete input schemas and descriptions. Naming follows verb_noun pattern (search_*, retrieval, count). Descriptions are detailed (100-250 chars avg) and explain WHAT/WHEN/HOW. Parameters are well-typed with enums for constrained inputs (view, id_type, sort). However, output schemas are NOT documented, LLMs cannot see what fields to expect from responses. No error handling guidance (recovery steps, retryability classification). No tool annotations (readOnlyHint, idempotentHint). Parameter descriptions lack format constraints (e.g., 'Scopus ID' but no example format). Missing dependency hints between tools (e.g., 'Call scopus_search first to get scopus_id for abstract_retrieval'). All tools are READ_ONLY but this is not declared in annotations.
Retrieve detailed metadata and abstract for a specific Scopus document by its identifier. Supports lookup by Scopus ID, EID, DOI, PII, or PubMed ID.
Retrieve detailed information about a specific academic institution or research affiliation from Scopus by affiliation ID or EID. Includes document counts, author lists, and address details.
Search for academic institutions and research affiliations in Scopus. Returns affiliation names, locations, document counts, and identifiers. Use field codes: AFFIL(), AF-ID(), CITY(), COUNTRY().
Retrieve full-text or metadata of a ScienceDirect article by DOI, PII, or EID. Access depends on your API key entitlements.
Retrieve a detailed author profile from Scopus by author ID, EID, or ORCID. Includes publication metrics, subject areas, and affiliation history.
Output schemas not documented. LLMs cannot see what fields responses contain, forcing them to guess downstream field names and risk broken tool chains.
No tool annotations. All 14 tools are READ_ONLY but lack readOnlyHint annotation. Agents cannot distinguish safe tools from destructive ones.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 60 | 2026-07-28+ | v2 |
Search for authors in the Scopus database. Returns author profiles including name, affiliation, document count, and identifiers. Use field codes: AUTHLASTNAME(), AUTHFIRST(), AFFIL(), ORCID().
Get citation counts for one or more Scopus documents. Supports batch lookup by providing comma-separated identifiers. Returns total citation count for each document.
Get a detailed citation overview for a Scopus document, including yearly citation breakdowns. Can filter by date range, exclude self-citations, and exclude book citations.
Retrieve the full text of a ScienceDirect article as clean plain text. Requires institutional access or entitlements for most articles.
Retrieve PlumX Metrics for a scholarly document. Returns aggregate altmetric counts including citations, usage, captures, mentions, and social media activity.
Search the Scopus database of abstracts and citations. Returns metadata about scholarly documents including titles, authors, DOIs, citation counts. Use Scopus field codes in queries: TITLE(), AUTH(), AFFIL(), DOI(), KEY(), PUBYEAR, etc.
Retrieve detailed information about a specific journal or serial publication by its ISSN. Includes publisher, subject areas, and CiteScore metrics.
Search for academic journals and serial publications in Scopus by title, ISSN, publisher, subject area, or open access status.
Retrieve subject area classifications used in Scopus or ScienceDirect. Useful for finding subject codes to filter searches. Returns classification codes, abbreviations, and descriptions.
No error handling guidance. Responses lack recovery hints (e.g., 'Try search_users() if lookup fails'). Errors are not classified as retryable vs user-fixable.
Parameter descriptions lack format constraints. 'Scopus ID' and 'DOI' lack examples or regex patterns. LLMs cannot validate input format before calling.
No dependency hints between tools. Descriptions do not explain when to call search_* before retrieval_*, forcing agents to discover chains via trial-and-error.