Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
Python agentaianthropicclaude-codecompressioncontext-engineeringcontext-windowcursorfastapilangchainllmmcpopenaiprompt-engineeringproxypythonragtoken-optimizationtokenstypescript
fix(mcp): pin mcp dependency to <2.0.0 to prevent server startup crash (#2642)
Robert Schorrb3f016b86637
diff --git a/pyproject.toml b/pyproject.tomlindex 1a43c32c5..6e1b1ba48 100644--- a/pyproject.toml+++ b/pyproject.toml@@ -78,7 +78,7 @@ proxy = ["orjson>=3.9.14; platform_python_implementation != 'PyPy'","httpx[http2]>=0.24.0","openai>=2.14.0", # OpenAI API format support- "mcp>=1.28.1", # MCP server (headroom_compress, retrieve, stats)+ "mcp>=1.28.1,<2.0.0", # MCP server (headroom_compress, retrieve, stats)"magika>=0.6.0", # ML content detection for ContentRouter"zstandard>=0.20.0", # Decompress zstd request bodies (Codex, etc.)"websockets>=13.0", # WebSocket proxy for /v1/responses (Codex gpt-5.4+)@@ -225,7 +225,7 @@ autogen = []# MCP server for Claude Code integrationmcp = [- "mcp>=1.28.1",+ "mcp>=1.28.1,<2.0.0","httpx>=0.24.0","starlette>=0.27.0","uvicorn>=0.23.0,<1.0",