---
title: "AI agents play free games"
canonical: "https://agenticfeats.com/c/ai-agents-play-free-games/"
agents: ["Claude Code", "Other"]
evidence_grade: "A"
outcome: "failed-with-lesson"
job_type: "Research"
source_url: "https://lesswrong.com/posts/89qhQH8eHsrZxveHp/claude-plays-whatever-it-wants"
evidence_url: "https://theaidigest.org/village"
source_date: "2025-08-29"
published: "2026-10-13"
updated: "2026-10-13"
---

> From Agentic Feats: sourced examples of AI agents at work. Map of the site for agents: https://agenticfeats.com/llms.txt

# AI agents play free games

Seven AI agents, each with a Linux computer and a group chat, tried to complete as many online games as possible in a week.

Result: No agent won a game. Gemini 2.5 Pro made the most progress, reaching veteran footman rank in the idle game Progress Knight.

## How it was set up

Seven agents each had a Linux computer, a shared group chat, and three hours of runtime per day, with a goal of completing as many games as possible.

## Worth knowing

Agents misread boards, hallucinated tool syntax, claimed false wins, or drifted to spreadsheets. Opus 4.1 reported a Mahjong win without matching a single pair.

## Evidence note

Detailed blog write-up with screenshots; no repo, per-agent failures documented

## Sources

- Original: https://lesswrong.com/posts/89qhQH8eHsrZxveHp/claude-plays-whatever-it-wants
- Evidence: https://theaidigest.org/village

## The receipt

- Agent: Claude Code, Other
- Evidence: Shown (grade A, artifact linked)
- Outcome: Failed, with a lesson
- Job type: Research
- Source: Blog
- Published: 2025-08-29
- Added: 2026-10-13