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 all mypy type errors and flaky embedding test
chopratejas3ab0a5dba030
diff --git a/headroom/ccr/mcp_server.py b/headroom/ccr/mcp_server.pyindex 09c049d5c..fe21d0074 100644--- a/headroom/ccr/mcp_server.py+++ b/headroom/ccr/mcp_server.py@@ -53,7 +53,7 @@ try:HTTPX_AVAILABLE = Trueexcept ImportError:HTTPX_AVAILABLE = False- httpx = None+ httpx = None # type: ignore[assignment]from .tool_injection import CCR_TOOL_NAMEdiff --git a/headroom/integrations/langchain.py b/headroom/integrations/langchain.pyindex 6ad0ea5c4..38afb9af3 100644--- a/headroom/integrations/langchain.py+++ b/headroom/integrations/langchain.py@@ -55,8 +55,8 @@ except ImportError:LANGCHAIN_AVAILABLE = FalseBaseChatModel = objectBaseCallbackHandler = object- Field = lambda **kwargs: None # noqa: E731- PrivateAttr = lambda **kwargs: None # noqa: E731+ Field = lambda **kwargs: None # type: ignore[assignment] # noqa: E731+ PrivateAttr = lambda **kwargs: None # type: ignore[assignment] # noqa: E731from headroom import HeadroomConfig, HeadroomModefrom headroom.providers import OpenAIProvider