Show HN: Scry, programmable internet search w/ congestion pricing
summary
Scry is a Model Context Protocol (MCP) server that allows AI agents to run complex, Turing-complete SQL-like queries over billions of internet documents from sources like Reddit, Hacker News, and arXiv. Unlike traditional search, which returns a handful of links, Scry enables agents to perform large-scale data analysis, graph walks, and vector-based semantic searches across a massive, structured public record. The system aims to provide a high-performance, traceable infrastructure for non-commercial research and commercial data engagement.
key points
- The platform transforms the web into a queryable database, allowing agents to perform complex operations such as recursive thread analysis, set intersections of citations, and time-series tracking of specific phrases.
- Scry integrates advanced vector operations directly into its query language, enabling users to define concepts via centroids or contrast axes and rank entire corpora based on these semantic directions.
- The infrastructure is designed for high-speed execution and accessibility, integrating with popular AI clients like Claude and ChatGPT via MCP to provide agents with real-time access to hundreds of billions of rows of data.
who should read it
AI developers and researchers who need to perform large-scale, programmatic analysis of public web data without building their own crawlers or indexes.
discussion positive
consensus
Commenters generally find the search tool technically impressive and valuable, but they are frustrated by the website's user interface and confusing pricing terminology. Many users find the homepage text to be unreadable or LLM-generated, leading to requests for a simpler, more human-centric layout. Despite the aesthetic complaints, there is a strong sense of curiosity and support for the tool's potential to challenge dominant search engines.
top insight
A user suggests distributing datasets via peer-to-peer torrents, similar to Wikipedia's snapshot method. This approach would democratize access to increasingly walled-off data while simultaneously reducing server congestion caused by users extracting large amounts of data through the query interface.
Generated by an LLM from the article and thread. Check the sources before relying on it.