Search and synthesize academic papers from arXiv, Semantic Scholar, Zotero, and local PDF libraries.
npx skills add https://github.com/zjunlp/Mechanist --skill research-lit-zjunlp
Please help me install this Agent Skill. Skill: research-lit Source: https://github.com/zjunlp/Mechanist/tree/main/skills/research-lit Command: npx skills add https://github.com/zjunlp/Mechanist --skill research-lit-zjunlp
High-intent search queries and answers about installing and using this skill.
Use the arxiv_fetch.py script with a search subcommand and a query string, optionally limiting results with --max. It calls the arXiv API and returns structured JSON metadata including title, abstract, authors, and categories.
Use the semantic_scholar_fetch.py search command with filters like --fields-of-study and --publication-types to surface IEEE, ACM, and Springer venue papers. It returns citation counts, venue metadata, and TLDR summaries that arXiv-only search misses.
Yes, the workflow degrades gracefully. If a Zotero or Obsidian MCP server is unavailable, that source is skipped silently and the remaining base sources — local PDFs, web search, arXiv, and the mechanic-db cloud search — still run.
RESEARCH_LIT.md is a raw audit dump of every retrieved paper with verbatim abstracts, intended only for human inspection. LANDSCAPE.md is the synthesized report with a structured paper table, narrative summary, and structural gaps that downstream skills read.
Unauthenticated requests to the Semantic Scholar Graph API throttle aggressively with HTTP 429 responses. Set the SEMANTIC_SCHOLAR_API_KEY environment variable for higher rate limits; the script retries 429 and 5xx responses with backoff.
satsuki-64Identify and synthesize research papers from Zotero, Obsidian, PDFs, and web sources.
mobled37Automate academic research pipelines from literature review to experiment planning.
OpenLAIRMaintains a persistent research knowledge base of papers, ideas, experiments, and claims.
jpologAutomate scholarly literature discovery across multiple databases with reproducible logging.
