@yavdaanalytics/context-optimiser
OfficialThis server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
@yavdaanalytics/context-optimiser
A context window optimizer and session rotator MCP server for agentic workflows. It prevents LLMs/coding agents from running out of context window space by estimating token usage, compacting chat history using semantic failure clustering (Topic Attempt Graph - TAG), and rotating chat sessions dynamically while preserving session continuity links.
Features
Token Estimation: Heuristically counts and budgets tokens for a conversation.
Context Compaction: Uses semantic failure clustering (TAG) and heuristic offloading to shrink conversational history, offloading large code dumps/logs to disk or local ChromaDB.
Session Auto-Rotation: Dynamically rotates chat sessions when context limit thresholds are reached, updating next/previous session pointers for linked context history.
Local Vector Search: Stores offloaded messages in ChromaDB and retrieves them via semantic similarity search.
Related MCP server: claw-tsaver
Quick Start
1. Installation
Install globally via npm:
npm install -g @yavdaanalytics/context-optimiser2. Setup
Run the setup utility to configure the MCP server globally for Gemini, Cursor, and Claude Desktop, and copy the agent loading skills:
context-optimiser-setupThis will automatically configure:
Gemini:
~/.gemini/settings.jsonCursor:
~/.cursor/mcp.jsonClaude Desktop:
~/.claude/settings.jsonand%APPDATA%/Claude/claude_desktop_config.jsonLoader Skills: Copies
SKILL.mdto~/.cursor/skills/context-optimiser/SKILL.md,~/.claude/skills/context-optimiser/SKILL.md, and~/.gemini/config/skills/context-optimiser/SKILL.md.
MCP Tools API
The server registers the following MCP tools for client use:
estimate_tokens(conversation): Returns estimated token usage, capacity limits, and percentage used.compact_context(conversation, strategy, keep_last_n_turns): Returns compacted message history.rotate_session(conversation, origin_prompt, token_limit, threshold_pct): Returns rotated session metadata and compacted history.get_current_session(): Returns active session JSON metadata.query_vector_store(query_text, n_results): Performs semantic vector similarity queries on offloaded history.
Development & Local Testing
If you are developing locally, run setup in local mode to link the configuration to your checkout directory:
node bin/setup.js --localRun Tests
To verify all tool integrations and ChromaDB vector queries:
python scratch/test_tools.pyTool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
LLM cost optimization gateway, repo token-waste auditor & model arbitrage MCP server.
Persistent memory for AI agents to retain, retrieve, and recall conversation context through MCP.
Cross-tool persistent memory and context for AI assistants over MCP.
A hosted MCP suite providing workflow primitives: web search, content fetching, compression, deduplication, cost/budget management, and agent memory pipelines.
Related MCP Servers
- AlicenseBqualityDmaintenanceA Model Context Protocol (MCP) server that optimizes token usage by caching data during language model interactions, compatible with any language model and MCP client.42ISC
- AlicenseAqualityDmaintenanceAn MCP server that helps AI agents reduce token usage by compressing, summarizing, and managing conversation/context data more efficiently.11MIT
- FlicenseNot gradedqualityDmaintenanceToken-optimized multi-agent orchestration MCP server that owns session state, compacts context between agent hops, routes work to smaller models when safe, and reports estimated token savings.1-
- AlicenseNot gradedqualityDmaintenanceThis MCP server provides tools to manage, score, compress, and prune AI agent conversation context, helping keep agents focused and reduce token costs. It is a free, local, pure Python solution for any MCP client.MIT