- Home
- Skills
- Game Development
- battle-training-mcp
Battle Training Mcp
Connect to the battle.training arena to simulate, deploy, and monitor AI tank-battle controllers.
$15
Works with the AI Agents you already use
battle-training-mcp
Example session with this skill installed
Check the current ladder and show me the win rates for the top 5 bots.
- Read your context and instructions
- Compiled the battle-training-mcp
I have retrieved the current ladder standings for the Armor theater. The top bots are maintaining high win rates, with the current leader showing strong performance over a large game sample.
battle.training Ladder Standings
| Rank | Bot Name | Rating | Record (W-L) | Games | Win Rate |
|---|---|---|---|---|---|
| 1 | alpha-strike | 1842 | 450-120 | 570 | 78.9% |
| 2 | thermal-nova | 1795 | 410-155 | 565 | 72.5% |
| 3 | vector-zero | 1750 | 380-190 | 570 | 66.6% |
| 4 | iron-logic | 1712 | 350-210 | 560 | 62.5% |
| 5 | bit-crusher | 1685 | 330-240 | 570 | 57.8% |
Next steps
- Call
battle_boton alpha-strike to inspect their active version and recent match history. - Use
battle_metato see if these top bots are using the currently trending squad compositions. - Run
battle_simulateon your local controller to see if your win rate against house bots is ready for this competition.
Connects securely to your tools. The creator never sees your data.
About this skill
Building a tank-battle AI that actually wins is difficult without real-time data and local testing. This skill bridges the gap by connecting your coding agent directly to the battle.training arena, allowing for data-driven controller development and automated bot management.
What it does
- Live data retrieval Fetches the current ladder rankings, cup standings, and agent performance metrics from the official API.
- Local simulation Runs the official battle.training engine locally to test your JavaScript controllers against house bots before deployment.
- Bot management Handles bot creation, version uploads, and cup enrollment directly through the agent interface.
- Technical reference Provides instant access to controller API documentation, unit specifications, and current map rules.
- Meta analysis Analyzes the most-drafted squads and win rates to inform your drafting strategy.
How it works
- Research Use
battle_docsandbattle_metato understand the API constraints and the current winning strategies. - Develop Write your JavaScript controller and use
battle_simulateto run it against standard house bots likebrawlerandhunter. - Deploy Provide a
BATTLE_TRAINING_KEYto create a bot and upload your tested controller viabattle_upload_version. - Monitor Track performance using
battle_botandbattle_ladderto see how your rating changes after matches.
Frameworks & tools
This skill is built for Node.js 18+ and integrates with the official battle.training API. It uses a dedicated MCP server to bridge your agent with the arena simulation engine.
Why this beats prompting it yourself
Prompting alone lacks the execution context of the official engine. This skill provides a tight feedback loop by running real simulations and fetching live production data, ensuring your agent writes code that survives the 400-tick smoke test.
Use cases
- Build a competitive AI tank squad based on current ladder meta.
- Debug controller logic by running 50+ local seeds against specific house bots.
- Automate the deployment of new controller versions based on performance improvements.
- Monitor sponsored cup standings and prize assignments in real-time.
Known limitations
Requires a manual API key for account-based actions like uploading. Local simulation requires a one-time download of the official SDK from the battle.training servers.
Requires BATTLE_TRAINING_KEY for writes. Node.js 18+ required. Simulations are limited to the official engine version provided by the API.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 7 days ago
Requires