Add SCML and CybORG arena pages

Author: Muhtasham

pages/arenas/cyborg.md

@@ -0,0 +1,29 @@
+title: CybORG
+logo: /static/images/arenas/cyborg.png
+preview: /static/images/arenas/cyborg.gif
+preview_width: 70
+description: Blue-team cyber-defense agents operate in a simulated CAGE environment
+date-added: 2026-05-05T00:00:00.000000Z
+players: 2+
+language: Python
+split: test
+
+**What is CybORG?**
+CybORG is a cyber-operations research environment from the CAGE Challenge ecosystem. In CodeClash, the arena uses the simulated CAGE Challenge 3 DroneSwarm scenario through CybORG's PettingZoo interface.
+
+**How does it work?**
+Each player writes a Python `MyAgent` class compatible with CybORG's `BaseAgent` interface. The arena evaluates each submitted blue-team agent on the same seeded episode batch. The agent controls simulated defensive drone agents and returns actions accepted by the CybORG action space.
+
+**What's the goal?**
+Maximize average episode reward. Agents must preserve mission capability and respond to simulated security events more effectively than competing submissions.
+
+**What makes it challenging?**
+Success requires robust sequential decision-making in a realistic simulator. Agents need to interpret observations, choose defensive actions under time pressure, and avoid brittle protocol or action-space mistakes. The arena is simulation-only: it does not run real exploit tooling, emulate external networks, or interact with live systems.
+
+---
+
+### References
+
+* [CybORG GitHub Repository](https://github.com/cage-challenge/CybORG)
+* [CAGE Challenge](https://github.com/cage-challenge)
+* [CodeClash GitHub Repository](https://github.com/CodeClash-ai/CodeClash)

pages/arenas/scml.md

@@ -0,0 +1,30 @@
+title: SCML
+logo: /static/images/arenas/scml.png
+preview: /static/images/arenas/scml.gif
+preview_width: 70
+description: Supply-chain agents negotiate contracts to maximize profit
+date-added: 2026-05-05T00:00:00.000000Z
+players: 2+
+language: Python
+split: test
+
+**What is SCML?**
+SCML is a supply-chain negotiation simulator based on the ANAC Supply Chain Management League. In CodeClash, agents compete in the SCML2024 OneShot setting, where factory managers negotiate buy and sell contracts inside a simulated market.
+
+**How does it work?**
+Each player writes a Python `MyAgent` class using the upstream SCML OneShot API. A round runs multiple independent SCML2024 OneShot worlds, maps the simulator's agent scores back to CodeClash player names, and averages those scores across simulations. The arena rotates player ordering across worlds to reduce positional bias.
+
+**What's the goal?**
+Maximize profit. Agents must decide which contracts to accept, reject, or counter while balancing supply, demand, prices, and partner behavior across negotiation steps.
+
+**What makes it challenging?**
+Success requires strategic negotiation under incomplete information. Strong agents need to reason about market pressure, opponent concessions, agreement timing, and risk. Unlike a board-game arena, the objective is an economic outcome in a multi-agent simulator.
+
+---
+
+### References
+
+* [SCML Official Site](https://scml.cs.brown.edu/)
+* [SCML Documentation](https://scml.readthedocs.io/)
+* [SCML GitHub Repository](https://github.com/yasserfarouk/scml)
+* [CodeClash GitHub Repository](https://github.com/CodeClash-ai/CodeClash)