Tech • IA • Crypto
Agent Battle: Mine the most diamonds in 45 minutes
TL;DR
A technical workshop pits artificial intelligence agents against each other in an optimization race to extract as many diamonds as possible in Minecraft within a limited time.
KEY POINTS
A real-time agent competition
Participants were invited to design and fine-tune agents capable of mining diamonds within 35 minutes, with each attempt lasting 5 minutes. The goal was simple on the surface: obtain the highest number of diamonds, with a live leaderboard adding competitive pressure.
Three major technical objectives
The exercise aimed to demonstrate the creation of a managed agent, the impact of configuration choices, and the importance of continuous improvement through evaluations. Participants had to adjust key parameters such as the system prompt, the model used, and integrated tools to influence agent behavior.
A standardized environment for performance comparison
All competitors started with the same seed, the same initial equipment, and an identical environment, ensuring fair comparison. Each agent operated in a Minecraft clone connected to a Mind Flare bot, without a direct visual interface, but with structured commands such as moving or interacting with blocks.
Optimization levers focused on code and strategy
Participants mainly worked in a my_agent.py file, where they could modify the model, write a system prompt, and integrate specific skills. The use of MCP servers and preconfigured modules helped accelerate testing while still allowing room for experimentation.
The importance of rapid evaluations
A fast evaluation system of about one minute enabled efficient iteration. This approach, described as “hill climbing,” involves continuously testing, measuring, and adjusting to improve agent performance within a limited timeframe.
A constraint on token efficiency
Beyond the number of diamonds, a decisive criterion was the diamonds/tokens ratio. In case of a tie, the agent that used fewer computational resources was favored, pushing participants to prioritize fine optimization rather than relying on heavier models.
Close results and unexpected performances
Near the end, several participants were tied, with an observed ceiling around 19 diamonds. One competitor ultimately surpassed this score in the final minutes, illustrating the impact of late-stage optimization. However, an anomaly was noted with a score showing 0 tokens, calling some intermediate rankings into question.
Technical constraints in real-world conditions
Connection issues related to Wi-Fi disrupted some participants, particularly involving Cloudflare, highlighting the practical challenges of deploying agents in shared, high-load environments.
CONCLUSION
This competition highlights the rise of autonomous agents and the importance of fine-tuned configuration, where raw performance and resource efficiency become inseparable.
Full transcript
What's going on everyone? Let's get settled in. Uh, this is going to be an agent battle and we're going to be tight on time. So, we're gonna get right to it. My name is Ben. I'm accompanied here by Jeff. We're on the applied AI team here at Anthropic, which means we help folks like yourselves squeeze the most juice out of the cloud ecosystem. Uh, and today we're going to be mining for diamonds in Minecraft. So, you might have played Minecraft. Uh, I did in the 2000s, 2010s, but in the 2020s, we have agents play Minecraft for us. And that's what we're going to be doing today. So, what are we actually going to accomplish here today? There's really three things. Number one, we're going to learn how to build and deploy a managed agent. This is something that we released a few weeks ago. Uh it's hosted on our infrastructure and we've set up a lot of the configuration for you, but your job is going to be to get the agent into peak diamond mining condition. Number two is we're going to understand the impact of an agent configuration. That's the system prompt, the model string, the skills and MCPs that you're plugging into your agent to get it to elicit the behavior that you're actually looking for. And then number three, you heard Will talk about this in the last workshop session if you were here. We're going to learn how to hill climb on evals. This is how we improve all of our agents internally. If you've deployed one into prod, you understand the process of measuring its impact, understanding its behavior, and then making changes to iteratively improve. And this is an agent battle, but there are some rules to the battle. Uh, number one is we're going to have a timer for about 35 minutes in which you're going to be able to build and experiment with your agents. Uh, you can only submit uh well, we're only going to accept one run per person. So, you can submit as many as you'd like, but we're only taking your top run. Each run is five minutes of the agent mining for diamonds. You can kill the run at any point with just like a quick control C in the terminal. Um, we have an eval set that it takes approximately one minute if you just want to quickly iterate on the agents and whoever has the most diamonds at the end of 35 minutes wins. We're going to have a leaderboard up here going to show you where you're at. There's also going to be a chat where your agents can talk to one another. Uh, and if anyone is in a tie at the end of the workshop, uh, we're going to be settling that tie based on token efficiency. So, this is not just mine the most diamonds. it is get the best diamonds to tokens ratio. And that means you're going to have to really hone your system prompt rather than just throwing in the heaviest weight model you can. Now, I'm going to hand it over to Jeff to talk through the logistics of how the harness actually works and what you guys are going to be doing during the workshop. >> Hello. Uh, okay. So, the harness, uh, we're actually shipping quite a bit for you to get started. Um, what you're going to primarily be thinking about is how can I optimize my experience in trying to mine as many diamonds as possible. We're giving you a couple different tools to accomplish this. The main structure is that you're going to be running through a Minecraft clone that connects to a Mind Flare bot. If you're not familiar with Mind Flare, uh, you're not going to be relying on visuals. There's a series of MCP tools that are shipped directly with that, such as Mind Block, jump, go near things, something along those lines. don't have to think too much about this. The main levers that you're going to be focused on uh are about I think it's on the next slide, but uh basically along the lines of how do I optimize this run? Um everyone's going to start from the same seed. So there's no real optimization that needs to happen there. And every time you reset, you'll have the same start kit, the same seed to go with that. Um okay, so the the agent that you'll be operating out of, so there's a my agent.py that's included in the repo. Uh so if you go to slas aent battle you'll have access to the repo and that's where you'll be running this. Uh my Asian.pay is where everything should really be taking place. You can adjust the model, you can adjust the system prompt which is currently empty and then you have the opportunity to use a skill that's shipped directly from andropic or you can replace with your own skill. Uh and then we also are including an MCP server if you wish to adjust these things. So these are the things that or parameters that you have available to adjust. Uh so try things out. Run evals. Um we'll also be circulating around the room as well. So if you run into anything or want to uh discuss then feel free to do so. Um okay, yeah, this has largely already been covered. So I want to make sure that we have as much time as possible to get started. Um so we can actually move over to uh starting the countdown. Looks like several people have already joined. I'm going to reset the time now and you may begin. So, uh, these will all go away. Let me Okay, great. So, it's completely refreshed. Uh, and the time has begun now. So, feel free to get started. And if you have questions, we'll be around. It looks like we already have a run with 10. We have several other participants who are getting started. Um just to give slightly more context. So within the repo there's a uh cloud code skill that you can use to help with setup. Um if that is uh if anyone's having an issue then that should hopefully help with the process. A couple people announced that they were having issues with Cloudflare. I think it may be due to the fact that there's so many people trying to connect via the uh the conference Wi-Fi. Um it's a bit small, but I did put a command up here that found that it worked for at least one person. Um so feel free to give that a try if you're running into any issues. About five minutes remaining. We have a three-way tie. And for some reason, the first person is showing as zero tokens, which is highly suspicious. So, we may have a two-way tie. Unless you can exh seems like 19 might be the upper echelon of what's possible, at least so far. Have time for about one more run. We actually have somebody who's broken 19 with only one minute 20 seconds to go. Wow. You have to reveal your technique. 20 seconds. 10 9 8 7 6 5 4 3 2 one. Time's up. Um so thank you everybody for participating. Uh looks like we have a clear winner now and would love for everybody who was able to mine 19 diamonds uh to come find Ben and I. Uh because our second place has zero tokens, we don't know who actually won second and third place yet. So uh we'll invite everybody who is uh one through five to come up. We don't know what they'll win yet. You guys can smile in front of your >> Nice job everybody.