Search academic papers and extract research insights

Turns a coding agent into a literature-research assistant by orchestrating a bundled Python CLI (scripts/papers.py) that hits the free Semantic Scholar and arXiv APIs, downloads arXiv PDFs, and extracts their text with PyMuPDF. The agent decides which subcommand to invoke and how to combine results into a literature scan, a deep read of one paper, an impact analysis, or a reading list.

Use it when a coding agent needs to search academic literature, pull a specific paper's full text, assess a paper's impact, or assemble a reading list, grounded in real Semantic Scholar and arXiv data rather than the model's own recollection. It is the Skill-mode port of a companion MCP server by the same author - both share the same feature set, but this version ships as a Claude Code plugin installable with a single command, with no long-running MCP process to keep alive; the MCP version may fit better where a persistent MCP server is already the deployment model.

Input is a research query, a paper identifier or arXiv ID, or a request for a reading list. Output is the CLI's combined results assembled by the agent into one of four deliverables: a literature scan, a deep read of a single paper with its downloaded PDF text extracted via PyMuPDF, an impact analysis, or a reading list.

Built on the free Semantic Scholar API and the arXiv API for search and metadata, PyMuPDF for PDF text extraction after downloading arXiv papers, and a bundled Python CLI the agent drives directly - no separate MCP server process required, unlike the companion papers-mcp project.

Researchers and engineers who want an agent to search, read, and analyze academic papers grounded in real Semantic Scholar and arXiv data, installed as a single-command Claude Code plugin rather than a separate MCP server.