From fe312bedd06cffa0936ee82b04b031940b2ab02b Mon Sep 17 00:00:00 2001
From: reacher-z <mtrxcop@gmail.com>
Date: Mon, 27 Jul 2026 17:46:55 +0800
Subject: [PATCH] docs: mention ClawBench as a related web-agent benchmark

---
 README.md | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/README.md b/README.md
index d4b26ca..ceebdcb 100644
--- a/README.md
+++ b/README.md
@@ -328,7 +328,11 @@ python show_azure.py --json_config "experiments.json" --result_dir <path_to_down
 ## 🤖 BYOA: Bring Your Own Agent
 Want to test your own agents in Windows Agent Arena? You can use our default agent as a template and create your own folder under `src/win-arena-container/client/mm_agents`. You just need to make sure that your `agent.py` file features `predict()` and `reset()` functions. For more information on agent development check out the [BYOA Doc](./docs/Develop-Agent.md).
 
-## 👩‍💻 Open-source contributions
+## Related external benchmarks
+
+[ClawBench](https://github.com/TIGER-AI-Lab/ClawBench) is a complementary benchmark for AI web agents performing real-world tasks across live websites. It is independent of Windows Agent Arena and is not integrated or supported by this repository; the link is provided for readers comparing agent-evaluation environments.
+
+## 👩‍💻 Open-source contributions
 
 We welcome contributions to the Windows Agent Arena project. In particular, we welcome:
 - New open-sourced agents to be added to the benchmark