About this skill

Core Pain Points in Literature Retrieval

Engineers and researchers often encounter:
* Fragmented Resources: Chinese and English literature scattered across different databases with varying access requirements and query syntaxes.
* Difficult Chinese Access: Massive Chinese journal literature locked behind paywalls, with cumbersome full-text download processes.
* Quality Filtering Overload: Sifting through vast result sets to identify high-quality, highly relevant papers is time-consuming.

The Global Literature Knowledge Base skill addresses these via a unified API, primarily powered by the SmartLib Open Platform. It covers 1.228 billion metadata records and offers differentiated access paths.

Core Capabilities & Key Workflow

How the Skill Works

The skill acts as an API orchestration layer with a key workflow:

Credential & Quota Management:
- Mandatory Prerequisite: First-time users must provide an email for registration (/register) to establish quota tracking and enable future upgrades.
- Token-Bound Calls: Before every billable API call (e.g., /search), you must first call /consume to obtain a single-use consume_token. This token (~60s TTL) is for anti-abuse and must be passed to the search endpoint.
- Transparent Quota: Monthly free tiers are provided (100 searches + 10 downloads). Upon exhaustion, an automated upgrade prompt appears. Quota is consumed per successful API call; failed calls do not deduct quota.
Intelligent Retrieval & Acquisition:
- Dual Dataset Routing: Automatically routes requests to API 1 (8M authorized Chinese journals, downloadable) or API 4 (1B+ global literature metadata) based on user intent.
- Adaptive Strategy: Employs broad, narrow, and balanced search strategies. For instance, a "literature review" intent triggers a broad strategy to maximize recall.
- Tiered Full-Text Download:
  - Chinese Journals: Direct PDF download via API 3 for authorized content.
  - Foreign OA Literature: Automatically probes 10+ channels (ArXiv, Unpaywall, CORE, etc.). This process does not consume SmartLib quota.
  - Paywalled Literature: Provides metadata and source links only; full-text is unavailable.

Key Capabilities

Chinese Journal Search: 8M authorized articles with detail views and PDF downloads.
Global Literature Search: Covers papers, patents, standards, and theses, providing DOI, abstracts, funding info, and core journal indexing labels (SCI/EI/CSSCI etc.).
Source Links: Each record averages 4.75 original database links (e.g., Scopus, WoS) for authenticity verification.
Smart Expansion: Automatic synonym expansion and strategy suggestions (e.g., broader terms) when results are too few.

Capability Boundaries & Critical Notes

Not Supported (Key Constraints)

Paywalled English Full-Text: Global literature (API 4) returns metadata only. Full-text for closed-access papers is unavailable.
Bulk Export: No batch export to formats like EndNote or BibTeX.
Plagiarism Check or Novelty Search: Does not perform paper similarity checks or novelty analysis.

Critical Notes

Mandatory Rules: Never fabricate an email; always execute /consume before each search to get a token.
Quota Exhaustion: All search requests are rejected; the system automatically enters a paid upgrade flow.
Token Time-Sensitivity: consume_token is single-use with a short TTL and must be obtained on-demand.
Network Dependency: Requires an active network connection to the SmartLib API; offline use is impossible.

Screenshots

Use Cases

When writing a master's thesis or a journal article, you need to systematically search for domestic and international core journal literature in your field, obtain complete abstracts, reference lists, and download PDFs of key Chinese papers for detailed reading.
Before writing a literature review for a technical project, you need to conduct a broad search for global research progress, patents, and dissertations from the past five years on that topic, then categorize the results by core indexing status and theme.
Before finalizing a technical roadmap for an R&D project, you need to perform a preliminary novelty search to examine existing patent layouts and research outcomes in that direction both domestically and internationally.
When investigating a specific technology (e.g., 'mechanical properties of carbon fiber composites'), you need to simultaneously find experimental data papers in Chinese journals and frontier conference papers and patents in English.

Best For

Master's or PhD students writing a thesis or journal manuscript who need comprehensive, authoritative literature support and access to original texts.
R&D engineers and scientists responsible for technology scouting or patent analysis who need to quickly grasp the global research landscape and patent layout in a technical field.
Academic journal editors or peer reviewers who need to quickly verify the authenticity, source, and core indexing status of references.
Science & technology intelligence analysts or R&D managers who need to regularly track literature output trends, key institutions, and highly-cited papers in a specific technical domain.

Related Skills

AnySearch

@user_51511428/anysearch

A unified real-time search service supporting general web search, vertical domain queries, parallel batch search, and web page content extraction.

Knowledge Management 2026.08.30 ↓ 23,616

Knowledge Base Atomic Slicing Engine

@user_741dc82b/knowledge-engineering

A core ETL engine for RAG/knowledge base pipelines that decomposes long documents into semantically complete, retrieval-ready atomic slices.

Knowledge Management 2026.08.30 ↓ 19,429

Make-to-Markdown Smart Converter

@user_741dc82b/make-to-markdown

A smart converter that transforms various document formats into Markdown, with automatic dependency management, fallback conversion, and output cleaning.

Knowledge Management 2026.08.30 ↓ 18,964

Bidding RAG Data Assistant

@user_441e055b/shibo-bid-rag-dataset

A RAG-oriented assistant for bidding data, supporting retrieval, summarization, and structuring of bid announcements across all industries nationwide.

Knowledge Management 2026.08.30 ↓ 19,010