Merge branch 'main' into john/abclash

Author: john-b-yang

.env.example

@@ -11,3 +11,9 @@ OPENAI_API_KEY=
 GEMINI_API_KEY=
 XAI_API_KEY=
 DASHSCOPE_API_KEY=
+
+# Optional: Claude via a custom Anthropic-compatible endpoint (mini_anthropic_model.AnthropicModel).
+# Used by the *_llama.yaml ladder configs. Set all three to route Claude through your endpoint.
+LLAMA_API_KEY=
+LLAMA_BASE_URL=
+LLAMA_MODEL=

.gitignore

@@ -1,5 +1,8 @@
 # Project specific
 
+# macOS
+.DS_Store
+
 .vscode/**
 *.ipynb
 *.traj.json
@@ -230,3 +233,6 @@ docs/visualization/*.jsonl
 docs/visualization/*.png
 docs/visualization/code_org
 docs/visualization/elo_plots
+
+# Per-model llama endpoint configs — kept local only (obscures internal API details)
+configs/mini/models/

.pre-commit-config.yaml

@@ -1,7 +1,8 @@
 ci:
   autoupdate_commit_msg: "chore: update pre-commit hooks"
 
-exclude: '\.ipynb\.py$'
+# Skip notebooks-as-py and committed sample logs (raw game data + imported bot source).
+exclude: '\.ipynb\.py$|^logs/'
 
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks

CONTRIBUTING.md

@@ -101,13 +101,13 @@ CodeClash/
 │   ├── agents/          # AI agent implementations (MiniSWEAgent, etc.)
 │   ├── arenas/          # Game arena implementations
 │   ├── analysis/        # Post-tournament analysis tools
+|   ├── cli/             # CLI entrypoint
 │   ├── tournaments/     # Tournament orchestration
 │   ├── viewer/          # Web-based results viewer
 │   └── utils/           # Shared utilities
 ├── configs/             # Tournament configuration files
 ├── docs/                # Documentation (MkDocs)
-├── tests/               # Test suite
-└── main.py              # Main entry point
+└──  tests/               # Test suite
 ```
 
 ## Types of Contributions
@@ -170,7 +170,7 @@ Analysis tools live in `codeclash/analysis/`. We're particularly interested in:
 | Preview docs | `uv run mkdocs serve` |
 | Build wheel | `uv build --wheel` |
 | Build wheel + sdist | `uv build` |
-| Run a tournament | `uv run python main.py <config>` |
+| Run a tournament | `uv run codeclash run <config>` |
 | View results | `uv run python scripts/run_viewer.py` |
 
 ### Building Distributions

README.md

@@ -54,7 +54,7 @@ uv sync --extra dev
 cp .env.example .env  # Then edit .env with your GITHUB_TOKEN
 
 # Run a test battle
-uv run python main.py configs/test/battlesnake.yaml
+uv run codeclash run configs/test/battlesnake.yaml
 ```
 
 > [!TIP]
@@ -66,14 +66,14 @@ uv run python main.py configs/test/battlesnake.yaml
 
 ```bash
 pip install -e '.[dev]'
-python main.py configs/test/battlesnake.yaml
+codeclash run configs/test/battlesnake.yaml
 ```
 </details>
 
 Once this works, you should be set up to run a real tournament!
 To run *Claude Sonnet 4.5* against *o3* in a *BattleSnake* tournament with *5 rounds* and *1000 competition simulations* per round, run:
 ```bash
-uv run python main.py configs/examples/BattleSnake__claude-sonnet-4-5-20250929__o3__r5__s1000.yaml
+uv run codeclash run configs/examples/BattleSnake__claude-sonnet-4-5-20250929__o3__r5__s1000.yaml
 ```
 
 ## ⚔️ How It Works
@@ -82,7 +82,7 @@ uv run python main.py configs/examples/BattleSnake__claude-sonnet-4-5-20250929__
   <img src="docs/assets/flowchart.jpg" style="width: 70%" />
 </p>
 
-In CodeClash, 2+ LM agents compete in a **code arena** over the course of a multi-round tournament.
+In CodeClash, 2+ coding agents compete in a **code arena** over the course of a multi-round tournament.
 
 For the duration of the tournament, each agent is iteratively improving their own codebase to win a high-level, competitive objective (e.g., accumulate resources, survive the longest, etc).
 
@@ -100,18 +100,6 @@ The winner is the LM agent who wins the most rounds.
 CodeClash includes competitive programming games and simulation-backed arenas, including BattleSnake,
 Bomberland, CoreWar, CybORG, Halite, HuskyBench, RoboCode, RobotRumble, and SCML.
 
-Bomberland is a Bomberman-style grid arena based on Coder One's Bomberland competition. Agents edit
-a Python `bomberland_agent.py` implementation and compete to maximize average score across seeded
-simulations through survival, damage, kills, and destructible-block control.
-
-SCML is a supply-chain negotiation arena based on the ANAC Supply Chain Management League OneShot
-track. Agents edit a Python `scml_agent.py` implementation and compete to maximize average profit
-across multiple simulated supply-chain worlds.
-
-CybORG is a simulated cyber-defense arena based on the CAGE Challenge 3 DroneSwarm scenario. Agents
-edit a Python `cyborg_agent.py` implementation and compete to maximize blue-team reward across
-simulated episodes.
-
 ## 🚀 Get Involved
 
 - Check out our [docs](https://docs.codeclash.ai/) for more details on running different arenas, configuring tournaments, etc.