Skip to main content
Glama
benetche

ares-mcp

by benetche
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
ARES_MAILTONoContact email for the OpenAlex/Unpaywall polite pool. Resolved from --mailto, ARES_MAILTO, .env, git config, or placeholder ares-mcp@localhost.ares-mcp@localhost
ARES_MAX_PDF_BYTESNoMaximum PDF size in bytes (~50 MiB).52428800
ARES_PDF_TIMEOUT_SNoPDF download timeout in seconds.60
ARES_METADATA_TIMEOUT_SNoMetadata HTTP timeout in seconds.20

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
}
logging
{}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}
extensions
{
  "io.modelcontextprotocol/ui": {}
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
search_papersA

Search open catalogs (OpenAlex + arXiv) for academic papers.

Use this first when the user needs papers, citations, or a reading list. Does not download PDFs. Prefer hits with pdf_url for the next step.

Ranking: pass sort=cited_by when the user wants the most cited / famous / popular papers. Pass sort=relevance (default) for topical match. Aliases like "most cited" are accepted; unknown values fall back to relevance.

A doi or arxiv_id looks up that paper (query may be empty). If query itself is a DOI or arXiv id, it is treated as a lookup, not keywords. Identifier filters win over a keyword query.

cited_by ranks on OpenAlex citation counts. arXiv keyword hits are skipped in that mode because they have no citation counts.

Possible error_code values: source_unavailable (one or both catalogs failed; remaining hits may still be useful). Empty papers is still status=ok.

download_and_parse_paperA

Download an open-access PDF. Does not create Markdown files.

Accepts a DOI or arXiv id. Resolves legal OA copies via OpenAlex locations, Unpaywall (preprints / accepted manuscripts), arXiv/PMC, and a few curated public URLs (e.g. the Bitcoin whitepaper on bitcoin.org). Does not bypass paywalls: publisher HTML is skipped.

Status values: ok, not_found, no_open_pdf, download_failed, invalid_path.

get_bibtexA

Fetch the official Crossref/DOI BibTeX entry and append it locally.

Skips the write when the DOI or cite key is already in the file (status=duplicate). Failures leave the .bib unchanged.

Status values: added, duplicate, not_found, error, invalid_path.

extract_sectionsA

Return named sections from a local PDF using heading heuristics.

Not a trained parser: headings are matched case-insensitively with aliases (e.g. Methodology ≡ Methods ≡ Materiais and Methods; Abstract ≡ Resumo). Missing sections are listed in unmatched; they do not fail the tool.

Status values: ok, file_not_found, parse_error, invalid_path. Does not fetch URLs — pdf_path must already exist in the workspace.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.4/5.0

Scored across 4 tools

Disambiguation5/5

Each tool targets a distinct stage of an academic-paper workflow: searching catalogs, downloading/parsing an OA PDF, extracting sections from a local PDF, and fetching BibTeX. The descriptions explicitly call out boundaries (e.g., search does not download, extract_sections does not fetch URLs), leaving no meaningful overlap.

Naming Consistency5/5

All four tool names follow the same snake_case verb-first pattern: extract_*, search_*, download_and_parse_*, get_*. The only variation, download_and_parse_paper, still uses the same conventions and is clearly readable.

Tool Count5/5

Four tools is well-scoped for the stated research-assistant purpose. Each tool covers a necessary step in the paper search-to-citation pipeline without redundancy or bloat.

Completeness4/5

The core lifecycle is covered: discovery, open-access retrieval, section-level reading, and BibTeX citation handling. Minor gaps exist, such as no way to manage or catalog local PDFs and no BibTeX support for arXiv-only papers, but agents can work around these.

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources