Brings Claude Code's context compaction to any MCP host, enabling agents to gauge context pressure, summarize history, re-hydrate files, and persist rules across session boundaries.
Claim it to get a verified publisher badge, a free copy of our full audit findings, and direct contact for any high-priority issues we find.
Install from
M8ven verifies MCPs across every public registry — install directly from whichever one you prefer.
process.env. You'll be asked to provide them before it can run.COMPACTION_ALLOWED_ROOTS— cwd colon-separated roots for file re-hydrationCOMPACTION_AUTO— false auto-compact on ingest when pressure ≥ nowPct (store mode only)COMPACTION_EMBED_API_KEYCOMPACTION_EMBED_BASE_URL— LLM base URL OpenAI-compatible /embeddings endpointCOMPACTION_EMBED_MODEL— embeddings model for semantic recall (e.g. nomic-embed-text); enables embedCOMPACTION_HOOKS— path to hooks JSON (see examples/hooks.example.json)COMPACTION_HOOKS_ENABLED— true set false to disable all hook executionCOMPACTION_LLM_API_KEY— optional bearer tokenCOMPACTION_LLM_BASE_URL— "": "http://localhost:11434/v1",COMPACTION_LLM_HEADERS— JSON of extra request headers (Azure api-key, gateway/tenant headers); overrides BearerCOMPACTION_LLM_MODEL— qwen2.5-coder:14b summarizer modelCOMPACTION_MODE— passthrough passthrough (host owns history) \ store (server holds it)COMPACTION_RECALL_MODE— auto auto \ embed \ lexical — recall ranking strategyCOMPACTION_STATE_DIR— ~/.compaction-mcp/sessions session + ledger persistenceCOMPACTION_SUMMARIZER— direct direct \ sampling \ auto (§6 of spec)[](https://m8ven.ai/mcp/ink01101011-compaction-mcp-eydv9c)