Context
GenGO is one of my projects where *ACL papers are indexed with some structured data, TLDR summaries, and semantic vectors, our aim is to help researchers to more efficiently navigate through fast increasing research papers. Using PoolsideAI coding agent, I have vibe-coded a MCP server that grants coding agents an access to a backend paper database server to allow iterative and flexible exploration.
I have been generating quick literature search reports using PoolsideAI agent and this MCP server for myself. To showcase this tool, I link some of the reports here. Note that I did not validate the output report.
If you want to use this MCP yourself, you need an access to the paper database. you have two options for this.
- Get an access to the online server
- Contact me for the access credentials
- Host the paper database yourself by doing
- Using docker, start Qdrant databaset server (instruction)
- Download the paper database snapshot from here
- Upload the snapshot to your local Qdrant server. Visit your local Qdrant dashboard UI here where you can do that easily
Sample reports
- Why Reasoning Is Effective with LLMs — Literature Report
- ACL 2025 → 2026 Research Topic Trend Analysis
- Text Embedding Geometry & Input Length Effects Literature Report
- ACL vs EMNLP: Comparative Research Topic Trend Analysis (2024–2026)
- the conclusion is interesting. ACL seems more “empirical” than EMNLP from the report
- “ACL has evolved into a systems-focused conference”
- “EMNLP has evolved into a reasoning-focused conference”