rafim-dev/mcp-context-condenser

Overview

Token-slimming AST code outliner, log compressor, and context budget analyzer for AI coding agents such as Cursor, Claude, Cline, and Antigravity. It reduces token usage and LLM API costs by up to 85% by condensing code structure and logs before they reach the model.

Best for

Best for
Developers using AI coding agents who need to stretch context windows and cut API spend on large codebases.

Use cases

• Trim large code files into compact AST outlines before sending to an AI agent
• Compress verbose logs to stay within context window limits
• Monitor and analyze context budget usage across coding sessions

How to use

Tools exposed

• condense_source
• extract_symbol
• compress_log
• token_budget

Tested with

Claude Desktop, Cursor, Cline, Antigravity

Example client config

{\n  "mcpServers": {\n    "context-condenser": {\n      "command": "node",\n      "args": ["/ABSOLUTE/PATH/TO/mcp-context-condenser/dist/index.js"]\n    }\n  }\n}

Notes

Token-slimming AST code outliner, log compressor, and context budget analyzer for AI coding agents such as Cursor, Claude, Cline, and Antigravity. It reduces token usage and LLM API costs by up to 85% by condensing code structure and logs before they reach the model.

0 stars on GitHub. Last updated 2026-09-07. Licensed MIT.

Use cases

• Trim large code files into compact AST outlines before sending to an AI agent
• Compress verbose logs to stay within context window limits
• Monitor and analyze context budget usage across coding sessions

Pros

• Significant token and cost reduction, up to 85%
• Works across multiple popular AI coding agents
• Built in TypeScript, easy to integrate as an MCP server

Cons

• No stars or community traction yet, early stage
• AST outlining may lose some code detail for complex or dynamic code
• Requires setup and configuration for each agent environment

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.