Aer MCP – A Model Context Protocol Server for Academic Papers

Aer MCP is a lightweight, extensible MCP server that lets AI assistants query the AEA (Association for Economic Analysis) research repository using natural language. By turning a static collection of academic papers into an LLM‑friendly API, it removes the need for custom web scrapers or manual database queries. Developers can embed scholarly search directly into chatbots, research assistants, or data‑analysis pipelines, enabling instant access to up‑to‑date literature without leaving the AI environment.

The server exposes a single, well‑defined resource: search. When an LLM client sends a request such as “Find recent papers on behavioral economics from AEA,” the MCP translates that into an HTTP call to the AEA API, retrieves the matching metadata (title, authors, abstract, publication date), and returns it in a structured JSON payload. This pattern fits naturally into the MCP request/response model, allowing developers to plug Aer MCP into any Claude or other LLM workflow that supports external tool calls.

Key features include:

• Natural‑language search – The server accepts plain English queries, which are internally converted into AEA’s query language.
• Rich metadata output – Results contain titles, authors, abstracts, and links, providing enough context for the LLM to summarize or cite.
• Extensible architecture – Built on the same foundation as other academic MCP servers (Stata‑MCP, NBER‑MCP), it can be expanded to support additional data sources or advanced filtering.
• Developer friendly configuration – A simple JSON entry in the MCP client config points to the Aer MCP executable, making deployment straightforward.

Typical use cases span academia and industry:

• Research assistants: Students or researchers can ask an AI to pull the latest AEA papers on a topic and receive concise summaries.
• Policy analysis: Analysts can quickly gather evidence from peer‑reviewed literature to support decision making.
• Educational tools: Tutors can generate reading lists or quiz questions based on recent research.

Integration is seamless: once the MCP server is running, any LLM client that supports external tools can invoke it with a single tool call. The server’s lightweight nature ensures low latency, and its open‑source implementation encourages community contributions—whether adding new filters or supporting additional academic databases.

In short, Aer MCP transforms static scholarly repositories into dynamic, AI‑ready knowledge bases, giving developers a powerful yet simple way to enrich conversational agents with authoritative academic content.