{"id":"1dark134-arxiv-mcp-server","name":"arxiv-mcp-server","af_score":34.0,"security_score":51.8,"reliability_score":32.5,"what_it_does":"Provides an MCP-based server (and associated client CLI) for searching, retrieving, analyzing, and exporting arXiv paper metadata and related citation/trend insights in multiple formats (e.g., BibTeX/RIS/JSON/CSV/YAML), with a REST API and a web UI described in the README.","best_when":"You want local or self-hosted tooling for arXiv discovery and research assistance that plugs into MCP-based agent workflows, plus export for downstream use.","avoid_when":"You need a fully specified, production-grade interface contract (OpenAPI spec, explicit MCP server URL, concrete auth/rate-limit semantics) without performing additional verification.","last_evaluated":"2026-03-30T15:38:42.733822+00:00","has_mcp":true,"has_api":true,"auth_methods":["API keys","OAuth (mentioned)"],"has_free_tier":false,"known_gotchas":["Provided content is mostly README-level claims; concrete MCP tool schemas, result formats, and error semantics are not included in the provided excerpt","Pagination behavior is only broadly referenced; agents may need to verify cursor/offset semantics before relying on it","Auth and rate limit details are not specified here, so agent clients may fail or be throttled without proper backoff handling"],"error_quality":0.0}