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
chore(merge): resolve upstream main conflicts
Kayzoaf9af3d2b52d
diff --cc headroom/cli/wrap.pyindex 41b7321ff,6fbec6551..1ccedb849--- a/headroom/cli/wrap.py+++ b/headroom/cli/wrap.py@@@ -818,16 -799,16 +818,19 @@@ def _ensure_proxy# Start (or restart) the proxy with the requested flagsclick.echo(f" Starting Headroom proxy on port {port}...")try:- proc = helpers._start_proxy(- proc = _start_proxy(-- port,-- learn=learn,-- memory=memory,-- agent_type=agent_type,-- code_graph=code_graph,-- backend=backend,-- anyllm_provider=anyllm_provider,-- region=region,-- openai_api_url=openai_api_url,++ proc = cast(++ subprocess.Popen[Any],++ helpers._start_proxy(++ port,++ learn=learn,++ memory=memory,++ agent_type=agent_type,++ code_graph=code_graph,++ backend=backend,++ anyllm_provider=anyllm_provider,++ region=region,++ openai_api_url=openai_api_url,++ ),)click.echo(f" Proxy ready on http://127.0.0.1:{port}")return proc