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
Add INFO-level logging for WS /v1/responses connect and complete
chopratejas92f74db8e010
diff --git a/headroom/proxy/handlers/openai.py b/headroom/proxy/handlers/openai.pyindex f088cff4f..a2eb3be34 100644--- a/headroom/proxy/handlers/openai.py+++ b/headroom/proxy/handlers/openai.py@@ -1103,7 +1103,7 @@ class OpenAIHandlerMixin:tokens_saved = 0# --- Connect to upstream OpenAI WebSocket ---- logger.debug(f"[{request_id}] WS connecting to {upstream_url}")+ logger.info(f"[{request_id}] WS /v1/responses connecting to {upstream_url}")# Use ssl=True to let the websockets library handle SSL natively.# Manual ssl.create_default_context() + certifi doesn't load the@@ -1164,6 +1164,10 @@ class OpenAIHandlerMixin:return_exceptions=True,)+ logger.info(+ f"[{request_id}] WS /v1/responses completed (tokens_saved={tokens_saved})"+ )+except Exception as ws_err:if ws_connected:# WS was established but broke mid-stream — re-raise