{"slug": "show-hn-limma-mcp-for-total-recall-and-precision-in-data-retrieval-for-agents", "title": "Show HN: Limma – MCP for total recall and precision in data retrieval for agents", "summary": "Berend Graham and cofounder Shiv Oberai launched Limma, a single MCP server built over the past year to improve retrieval precision and recall for AI agents working over RAG pipelines, Graham posted on Hacker News. Limma maps and structures data points within unstructured data and federates structured data for addressability so agents retrieve the right information instead of searching through chunks directly, using self-hosted models primarily from Qwen to keep deployment on-premises and fast. Graham said the approach yields the lowest hallucination rate compared with targeting agent action verification alone, and a free monthly version is available on the company's website.", "body_md": "Hello, my name is Berend Graham. My cofounder Shiv Oberai and I have been working on Limma for a year now with the goal of building a proactive solution to the reliability issues that plague practical AI and agentic use in companies with RAG.\n\nLimma is a single MCP that exists for the sole purpose of ensuring that the right information is retrieved with every query and no information is skipped and or missed. This means total precision and recall. We've found that this approach leads to the lowest rate of hallucination vs. solely targeting agent action verification which has made tremendous strides in capability.\n\nWe do this by building a map and structuring all data-points within unstructured data as well as federating structured data for addressability so that agents don't need to search through chunks directly and instead only get the right information . We wanted this to be extremely easy to use so that anyone can set it up and try it within minutes so that the precision issues (especially with large quantities of data) that LLMs struggle with can finally be remedied. We use self-hosted models (primarily Qwen models as of current) so that the system can be deployed on one's own server if needed, and also to optimize for speed.\n\nWe have a free monthly version on our website if anyone would like to try it out and provide feedback!\n\nComments URL: [https://news.ycombinator.com/item?id=50254187](https://news.ycombinator.com/item?id=50254187)\n\nPoints: 2\n\n# Comments: 0", "url": "https://wpnews.pro/news/show-hn-limma-mcp-for-total-recall-and-precision-in-data-retrieval-for-agents", "canonical_source": "https://limma.ai", "published_at": "2026-10-27 02:50:43+00:00", "updated_at": "2026-10-27 03:11:33.148099+00:00", "lang": "en", "topics": ["ai-agents", "agent-protocols", "ai-tools", "large-language-models", "ai-safety"], "entities": ["Limma", "Berend Graham", "Shiv Oberai", "Qwen", "Hacker News"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/show-hn-limma-mcp-for-total-recall-and-precision-in-data-retrieval-for-agents", "markdown": "https://wpnews.pro/news/show-hn-limma-mcp-for-total-recall-and-precision-in-data-retrieval-for-agents.md", "text": "https://wpnews.pro/news/show-hn-limma-mcp-for-total-recall-and-precision-in-data-retrieval-for-agents.txt", "jsonld": "https://wpnews.pro/news/show-hn-limma-mcp-for-total-recall-and-precision-in-data-retrieval-for-agents.jsonld"}}