Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.2.0] - 2026-07-19

Added

• Implemented summarize_paper MCP tool with a multi-tiered LLM orchestration pipeline (client-side sampling via PluMCP, server-side fallbacks to Gemini, OpenAI, and Anthropic, and local structural parsing).
• Integrated JSoup and Apache PDFBox for HTML and PDF document text extraction.
• Centralized server-side LLM interactions into a dedicated llm_api module.
• Added comprehensive unit and integration tests covering document parsing and JSON-RPC tool callbacks.
• Added complete standard Clojure docstrings and structured observability with org.clojure/tools.logging and slf4j-simple.

[0.1.2] - 2026-07-19

Fixed

• Updated CircleCI Docker image to next-gen cimg/clojure:1.11-openjdk-21.0 (OpenJDK 21) to satisfy Java 21 Virtual Threads compilation requirements from the plumcp library.

[0.1.1] - 2026-07-19

Changed

• Upgraded dependencies to latest stable versions (plumcp to 0.2.2 and lucene to 10.5.0), establishing Java 21 as a baseline requirement due to downstream Virtual Thread support in the web server layer.

[0.1.0] - 2026-07-19

Added

• Renamed project from research-mcp to scholarly-mcp and published to coordinates org.clojars.punit-naik/scholarly-mcp.
• Integrated 6 major academic data providers: OpenAlex, Crossref, Semantic Scholar, Springer Nature, arXiv, and PubMed.
• Added 7 MCP tools: search_papers, get_citations, retrieve_full_text, compare_papers, build_literature_review, export_citations, and find_nature_papers_by_abstract.
• Created automated integration and unit test suite verifying all tools and API endpoints.
• Added Git pre-commit webhook formatting validation using cljstyle.
• Configured CI pipeline with next-gen CircleCI image running OpenJDK 21.0.
• Created LICENSE file under the Eclipse Public License 2.0 (EPL-2.0).