Agent♥︎Age
Catalog

LLMSlim memory

Official

by Thanatos9404 · Python

Long-term memory and context fitting for agents, with no LLM in the loop. Local SQLite.

The io.github.Thanatos9404/llmslim MCP server provides long-term memory and context fitting for agents. It operates without an LLM in the loop and uses a local SQLite database. The server is intended to support agent memory and context handling through stored, fitted information.

🛠️ Key Features

  • Long-term memory for agents
  • Context fitting
  • No LLM in the loop
  • Local SQLite storage

🚀 Use Cases

  • Supplying agents with retained long-term memory
  • Fitting or aligning retrieved context for agent tasks

⚡ Developer Benefits

  • Predictable local data storage via SQLite
  • Memory and context handling without requiring LLM execution inline

⚠️ Limitations

  • Description only specifies long-term memory and context fitting; no additional capabilities are provided.

Topics

cost-optimizationllmprompt-engineeringpythonragtoken-optimizationcontext-compressionllm-agentsllmlinguaprompt-compressionagent-memoryclaude-code-plugingemini-cli-extensionlong-term-memorymcpmcp-servermodel-context-protocolai-agentsai-memoryclaude-code