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(ci): match proxy formatting to Ruff 0.15.9
JerrettDavis710ef8d13af6
diff --git a/headroom/proxy/handlers/openai.py b/headroom/proxy/handlers/openai.pyindex 328ab2367..40150e77e 100644--- a/headroom/proxy/handlers/openai.py+++ b/headroom/proxy/handlers/openai.py@@ -583,7 +583,11 @@ class OpenAIHandlerMixin:# Cacheif self.cache and response.status_code == 200:await self.cache.set(- messages, model, response.content, dict(response.headers), tokens_saved+ messages,+ model,+ response.content,+ dict(response.headers),+ tokens_saved,)await self.metrics.record_request(@@ -1517,7 +1521,10 @@ class OpenAIHandlerMixin:return JSONResponse(status_code=400,content={- "error": {"message": f"Invalid JSON: {e}", "type": "invalid_request_error"}+ "error": {+ "message": f"Invalid JSON: {e}",+ "type": "invalid_request_error",+ }},)diff --git a/headroom/proxy/helpers.py b/headroom/proxy/helpers.pyindex fea2d8273..98adbe177 100644--- a/headroom/proxy/helpers.py+++ b/headroom/proxy/helpers.py@@ -122,7 +122,12 @@ def _get_rtk_stats() -> dict[str, Any] | None:except Exception:pass- return {"installed": True, "total_commands": 0, "tokens_saved": 0, "avg_savings_pct": 0.0}+ return {+ "installed": True,+ "total_commands": 0,+ "tokens_saved": 0,+ "avg_savings_pct": 0.0,+ }async def _read_request_json(request: Request) -> dict[str, Any]: