An MCP server for LETHE, a context management system for long-running LLM agents that archives large tool outputs externally and recalls them on demand to reduce context growth and token usage. It exposes archive, recall, and status operations for Claude Code and Codex.
From the registry: Context GC for LLM agents: offload large tool outputs and recall them to save tokens.

$ pip install "lethe-llm-context[mcp]"https://pypi.org/project/lethe-llm-context/'}],Please install the `lethe` MCP server into my current AI client (that's you).
Required prerequisites (do these first if not already done):
- **Install Python package with MCP extras** — Install the Lethe package and its MCP server entrypoint. Run: `pip install "lethe-llm-context[mcp]"` (https://pypi.org/project/lethe-llm-context/'}],)
Canonical MCP server config (stdio transport):
- command: `lethe-mcp`
- args: []
Note: README shows a stdio MCP server launched via the `lethe-mcp` command after installing the Python package with MCP extras. No MCP-specific environment variables are documented. `ANTHROPIC_API_KEY` appears only for a separate real Claude demo, not for running the MCP server itself.
Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.ANTHROPIC_API_KEYrequiredRequired for running the real Claude demo exampleAI orchestration with hive-mind swarms, neural networks, and 87 MCP tools for enterprise dev.
Persistent codebase knowledge graph. Survives session restarts and context compaction.
A powerful toolkit for coding, providing semantic retrieval and editing capabilities.