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
feat(config): add OpenAI API URL override option
Nicola Beghin7ecf94c88102
diff --git a/headroom/proxy/server.py b/headroom/proxy/server.pyindex 577faae7f..0e45b6a41 100644--- a/headroom/proxy/server.py+++ b/headroom/proxy/server.py@@ -169,6 +169,7 @@ class ProxyConfig:# Serverhost: str = "127.0.0.1"port: int = 8787+ openai_api_url: str | None = None # Custom OpenAI API URL override# Optimizationoptimize: bool = True@@ -760,6 +761,10 @@ class HeadroomProxy:def __init__(self, config: ProxyConfig):self.config = config+ # Override OPENAI_API_URL with config if set+ if config.openai_api_url:+ HeadroomProxy.OPENAI_API_URL = config.openai_api_url+# Initialize providersself.anthropic_provider = AnthropicProvider()self.openai_provider = OpenAIProvider()@@ -2562,6 +2567,7 @@ if __name__ == "__main__":# Serverparser.add_argument("--host", default="127.0.0.1")parser.add_argument("--port", type=int, default=8787)+ parser.add_argument("--openai-api-url", help=f"Custom OpenAI API URL (default: {HeadroomProxy.OPENAI_API_URL})")# Optimizationparser.add_argument("--no-optimize", action="store_true", help="Disable optimization")@@ -2659,6 +2665,7 @@ if __name__ == "__main__":config = ProxyConfig(host=_get_env_str("HEADROOM_HOST", args.host),port=_get_env_int("HEADROOM_PORT", args.port),+ openai_api_url=_get_env_str("OPENAI_TARGET_API_URL", args.openai_api_url),optimize=optimize,min_tokens_to_crush=_get_env_int("HEADROOM_MIN_TOKENS", args.min_tokens),max_items_after_crush=_get_env_int("HEADROOM_MAX_ITEMS", args.max_items),