BioCite-MCP
<p align="center">
<img src="logo.jpg" alt="BioCite-MCP Logo" width="300">
</p>
# BioCite-MCP 🧬
[](https://pypi.org/project/biocite-mcp/)
[](https://pypi.org/project/biocite-mcp/)
[](https://opensource.org/licenses/MIT)
[](https://modelcontextprotocol.io)
**BioCite-MCP** is an advanced Model Context Protocol (MCP) server that acts as a real-time bridge between LLMs and academic literature databases (Europe PMC & Crossref). It eliminates citation hallucinations by forcing AI models to retrieve verified, peer-reviewed paper metadata and real DOIs directly within their workflow.
---
## 🌟 Key Features
### 🔍 Literature Discovery
- **`search_literature`**: Query Europe PMC for real biological papers. Returns structured metadata including DOIs and abstracts.
- **`find_related_papers`**: Discovers semantically related research using Europe PMC's Recommendations engine with an automated Citations/References fallback.
### 📝 Citation & Formatting
- **`resolve_citation`**: Converts any DOI into publication-ready citation strings (APA or Nature style).
- **`export_bibtex`**: Retrieves professional BibTeX entries directly via Crossref content negotiation.
### 🛡️ Manuscript Auditing & Support
- **`audit_manuscript`**: Scans your text for DOIs to verify them and flags potential citations that lack DOIs.
- **`summarize_paper`**: Fetches abstracts and prepares high-quality summarization prompts optimized for LLMs like Claude.
- **`check_duplicate_citations`**: Uses fuzzy matching (`rapidfuzz`) to identify and group duplicate research in your lists.
### 📚 Integration
- **`push_to_zotero`**: Seamlessly add verified papers to your Zotero library via the Web API.
---
## 🚀 Installation
```bash
pip install biocite-mcp
```
*Note: For development, use `pip install -e .` in the repository root.*
---
## 🔧 Configuration
Add `biocite-mcp` to your MCP host configuration (e.g., `claude_desktop_config.json`):
```json
{
"mcpServers": {
"biocite-mcp": {
"command": "python",
"args": ["-m", "biocite_mcp"]
}
}
}
```
---
## 🛠️ Usage Examples
1. **Search**: "Find recent papers about DREB2A drought stress in tomato."
2. **Resolve**: "Format the citation for DOI 10.1093/jxb/erx393 in Nature style."
3. **Analyze**: "Audit this manuscript draft for citation accuracy: [Your Text Here]" *(Pro-tip: Use Claude's Filesystem MCP to read your manuscript file and pipe the content directly into this tool!)*
4. **Export**: "Give me the BibTeX for 10.1111/j.1365-313X.2006.02701.x"
---
## 📜 License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
---
**Developed by ZaEyAsa — Your Advanced Agentic Bio-Citation Assistant.**
TDQS
Scored across 8 tools
Each tool has a distinct and clearly defined purpose: scanning text for DOIs, checking duplicates, exporting citations, finding related papers, pushing to Zotero, resolving citations, searching literature, and summarizing papers. No overlap in functionality.
All tool names follow a consistent verb_noun pattern with lowercase and underscores, e.g., 'audit_manuscript', 'export_bibtex', 'search_literature'. The naming is predictable and easy to understand.
With 8 tools, the set is well-scoped for a citation management server. It covers essential operations without being overwhelming or too sparse.
The tools cover core citation workflows: searching, validation, export, metadata retrieval, integration with Zotero, and duplication detection. Minor gaps like batch export or import are absent but do not hinder primary use cases.