Agent Skill: PubMed Literature RAG & Citation Agent Skill
📌 Description
Vectorizes PubMed abstracts using PubMedBERT, builds FAISS indices, and instructs agents to synthesize clinical evidence with exact PMID citations.
🤖 Agent Execution Protocol
When an AI Agent is tasked with pubmed-rag-synthesizer:
- Input Validation: Verify that the required input files or coordinates are supplied.
- Environment Check: Ensure dependencies (
LangChain, PubMedBERT, FAISS, EUtils API) are installed. - Execution: Run the protocol pipeline snippet below.
- Output Generation: Produce actionable Markdown/JSON summaries with publication figures.
💻 Protocol Code Snippet
def build_pubmed_rag(pmid_list):
# Vector Search Pipeline
print(f"Indexing {len(pmid_list)} PubMed abstracts for RAG synthesis.")
📥 Input & Output Specifications
Input Contract
- Target Files: Valid input data matching domain formats.
- Parameters: Quality thresholds and cutoffs.
Output Contract
- Results Table: Structured summary dataframe or matrix.
- Visualization: Rendered SVG/PNG figures.
📄 License
Distributed under the MIT License. See LICENSE for details.