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
test(copilot): use the real httpx in the hooks test instead of a stub (#3968)
Tejas Chopra0126fba63f5a
diff --git a/tests/test_proxy_copilot_auth_hooks.py b/tests/test_proxy_copilot_auth_hooks.pyindex 7a37a82c7..4bf7f33cb 100644--- a/tests/test_proxy_copilot_auth_hooks.py+++ b/tests/test_proxy_copilot_auth_hooks.py@@ -13,7 +13,6 @@ ROOT = Path(__file__).resolve().parents[1]_ISOLATED_MODULE_NAMES = ("headroom.proxy","headroom.proxy.handlers",- "httpx","fastapi.responses","tests.headroom_proxy_handlers_openai","tests.headroom_proxy_handlers_streaming",@@ -42,12 +41,10 @@ def _load_handler_module(monkeypatch: pytest.MonkeyPatch, module_name: str, relahandlers_pkg.__path__ = [str(ROOT / "headroom" / "proxy" / "handlers")]monkeypatch.setitem(sys.modules, "headroom.proxy.handlers", handlers_pkg)- httpx_mod = types.ModuleType("httpx")- httpx_mod.ConnectError = type("ConnectError", (Exception,), {})- httpx_mod.ConnectTimeout = type("ConnectTimeout", (Exception,), {})- httpx_mod.PoolTimeout = type("PoolTimeout", (Exception,), {})- httpx_mod.ReadTimeout = type("ReadTimeout", (Exception,), {})- monkeypatch.setitem(sys.modules, "httpx", httpx_mod)+ # The real httpx, not a stub: it is a core dependency, and a partial stub+ # breaks whatever else imports httpx during a cold import (the CCR/MCP SSE+ # client needs httpx.Response). It also keeps the exception hierarchy the+ # proxy catches (TransportError and its subclasses) exactly as in production.responses_mod = types.ModuleType("fastapi.responses")