Skip to main content
Glama
Sign In
Matryoshka

Matryoshka

by andrewb6868
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Local

Related Servers

Alternatives to Matryoshka

No user-submitted related servers found.

    Related Servers

    • lattice-mcp

      alessiomiyazaki08
      F
      license
      Not graded
      quality
      A
      maintenance
      Enables token-efficient analysis of very large documents by letting an LLM issue constrained symbolic commands while results and context are stored server-side as compact handles and memos. This avoids loading full documents or large intermediate outputs into the model context.
      4 npm
      -

    • Matryoshka

      JerryNovikov048
      F
      license
      Not graded
      quality
      A
      maintenance
      Enables token-efficient analysis of documents far larger than an LLM context window by executing validated symbolic commands server-side and returning compact handles or memos instead of full text.
      4 npm
      -

    • lattice-mcp

      Alessiaw11
      A
      license
      Not graded
      quality
      A
      maintenance
      Enables token-efficient analysis of very large documents by executing constrained symbolic REPL commands against server-side handles, memory pads, and recursive queries rather than passing entire documents through the model context.
      4 npm
      MIT

    • lattice-mcp

      Richardson60512
      F
      license
      Not graded
      quality
      A
      maintenance
      Enables MCP clients to analyze documents far larger than their context window by having the model emit validated symbolic S-expression commands that are parsed and executed server-side against the loaded content. In-memory handles, recursive sub-queries, and server-side memos keep token usage low while preserving full data for on-demand expansion.
      4 npm
      -

    • Matryoshka

      Asli179
      F
      license
      Not graded
      quality
      A
      maintenance
      Enables analysis of documents far larger than an LLM's context window without vector databases or chunking, by having the model emit constrained symbolic S-expression commands that a logic engine parses, validates, and executes over the source text. Handle-based in-memory storage and server-side memo pads keep results and stashed context as compact stubs, cutting token usage dramatically while supporting recursive sub-queries over chunks.
      4 npm
      -

    • lattice-mcp

      niklasnikolaidis8756
      A
      license
      Not graded
      quality
      A
      maintenance
      Enables analysis of documents far larger than an LLM's context window by letting the model emit validated symbolic commands that execute server-side, returning compact handles rather than full result sets. It also lets agents stash context as server-side memos that can be expanded later, cutting token usage across long sessions without vector databases or chunking heuristics.
      4 npm
      MIT

    This server cannot be deployed

    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    Maintenance

    ActivityNo data
    ResponsivenessSyncing

    Resources