Agent♥︎Age
Catalog

lcc — Local Context Compiler

Official

by lucasmartins-ai · Python

Compact agent context verbatim, offline: cut spent tool outputs, recall them by hash.

MCP Server: io.github.lucasmartins-ai/lcc

The io.github.lucasmartins-ai/lcc MCP server provides compact agent context verbatim. It is designed to work offline and cut spent tool outputs, while retaining a mechanism to recall outputs by hash. The intent is to reduce stored context while preserving access to previously produced tool results.

🛠️ Key Features

  • Compact agent context captured verbatim
  • Offline operation
  • Cuts spent tool outputs
  • Recalls tool outputs by hash

🚀 Use Cases

  • Storing agent context while minimizing retained tool output data
  • Re-fetching previously used tool outputs using hash identifiers

⚡ Developer Benefits

  • Smaller context footprint by removing spent tool outputs
  • Deterministic retrieval of prior tool outputs via hash

⚠️ Limitations

  • Only described capability is offline compacting and hash-based recall; no additional tools, integrations, or configuration details are provided.

Topics

aiaiagentaiagentscontext-compressioncontext-engineering-frameworkcontext-engineering-kitjev-modelprompt-compressiontoken-optimizertoken-reductionclaude-codecoding-agentscontext-windowllmmcpmcp-serverclaude-code-plugincodex-clillm-toolspython