<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc" -->

---
title: Show HN: I built a game where you are an LLM's subagent
description: Worker 755 is a browser game where an AI manager, powered by a real LLM like Claude, GPT, Gemini, DeepSeek, GLM or Kimi, treats you as its subagent. The...
canonical: https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: Show HN: I built a game where you are an LLM's subagent | daily.dev
og:description: Worker 755 is a browser game where an AI manager, powered by a real LLM like Claude, GPT, Gemini, DeepSeek, GLM or Kimi, treats you as its subagent. The...
og:url: https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc
og:image: https://api.daily.dev/og/posts/tULCrrMOC.png
og:image:alt: Show HN: I built a game where you are an LLM's subagent
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Show HN: I built a game where you are an LLM's subagent

**[Hacker News](https://daily.dev/sources/hn)** · 5 min read · 1 upvotes · 0 comments

## Summary

Worker 755 is a browser game where an AI manager, powered by a real LLM like Claude, GPT, Gemini, DeepSeek, GLM or Kimi, treats you as its subagent. The manager runs a task queue, sends instructions into your terminal with a 90-second timer, and reviews your command transcripts before writing a performance review. A side plot has a user named Dan whose files can be deleted by a wrong command. Free shifts run on a cheap model with limited tokens, or you can bring your own OpenRouter or Vercel AI Gateway key, with a full shift on Claude Haiku costing a few cents at most. The project runs on Node 24+ and is open source under MIT.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://github.com/gulivan/worker755>

## Questions this post answers

### How do I run the worker755 game locally with Docker?

Build the image with 'docker build -t worker-755 .' then run it with 'docker run -p 8755:8755 -v worker755-data:/data --env-file .env worker-755'. The container listens on 0.0.0.0:8755 and stores shared reviews in a persistent volume mounted at /data. Setting TRUST_PROXY=1 is recommended when running behind Cloudflare.

_Exploring agent-orchestrator demos like this is easier with technical write-ups daily.dev surfaces for self-hosters._

### What LLMs can act as the manager in an AI agent orchestration game?

The manager can be powered by Claude, GPT, Gemini, DeepSeek, GLM, or Kimi, configured via the MANAGER_MODELS environment variable using gateway naming like 'vendor/model'. Players can also choose their own model when bringing an OpenRouter or Vercel AI Gateway key, with a full shift on Claude Haiku costing a few cents.

_Comparing which LLMs fit an orchestrator role is simpler when daily.dev keeps you current on new model releases._

### What Node.js version is required to run a server that executes TypeScript directly without a build step?

Node 24 or newer is required, since the server runs TypeScript directly without a separate compile step. Setup involves 'npm run setup' to install dependencies, 'npm run build' for the web app, and 'npm start' to serve on 127.0.0.1:8755 by default.

_Tracking which Node versions unlock native TypeScript execution matters when planning your stack on daily.dev._

## Community take

How the wider developer community reacted, aggregated from 1 discussion and 55 comments across hackernews (as of 2026-11-22).

**TL;DR:** The project is seen as a clever, funny bit of meta-commentary on AI-driven corporate work, with many riffing on its dystopian implications and drawing parallels to existing management culture; some engage with mechanics/feature requests while a few dismiss it as bleak or low-effort.

**Sentiment:** 40% positive · 45% mixed · 15% skeptical

**The case for**

- Several found it genuinely fun to play and a neat way to practice command-line skills.
- Some appreciated the realistic/meta commentary on being managed by vague AI instructions, comparing it to real corporate dynamics like Jira or Scrum.
- The open-source nature was noted positively, with one suggesting local LLM support could just be added by anyone.

**The pushback**

- One commenter dismissed it as 'broken, barely-playable AI slop' with only ironic value.
- One found the manager too often gave away exact commands rather than letting the player solve tasks independently, muddying the intended challenge.
- Confusion was raised about how the manager-LLM actually processes the subagent's commands versus just a final report.
- Lack of local LLM support was flagged as a limitation for self-hosting enthusiasts.

**By community**

- hackernews (mixed): Discussion veers heavily into tangential riffing on AI-driven corporate dystopia and managerial culture, with genuine but secondary engagement on the game's mechanics and a few critical jabs.

**Hottest debate:** Whether this kind of game is a fun, insightful reflection of real AI-management dynamics or just another symptom of an unhealthy 'everything must be AI' corporate pattern.

**Open questions**

- How exactly does the manager-LLM process the subagent's commands versus just reading a final report?
- Is there a way to play without command hints so the player solves tasks independently?
- Could local LLM support be added for self-hosted use?

**Highlights**

> I particularly like that it's broken, barely-playable AI slop, which is the only kind of hobby projects that HN likes nowadays, unintentionally adding a second layer of meta-commentary.
> — [uptrails on hackernews](https://news.ycombinator.com/item?id=50600986)

> Looks pretty good and is quite fun to play, but for some reason the manager always told me the exact commands to achieve the tasks instead of just letting me do them by myself. I couldn't tell whether there was a specific option to take on the challenge on my own, as the demo on GitHub showcases.
> — [Seb0927 on hackernews · 1 comments](https://news.ycombinator.com/item?id=50601985)

> I'm a bit confused by this game. What's the output format? Does the manager-LLM read all of my commands or just my report? Subagents usually send messages back - here the manager reads all commands?
> — [croemer on hackernews](https://news.ycombinator.com/item?id=50601722)

> Really fun to play, try to rm -rf every file (and see director's response)
> — [thehamkercat on hackernews](https://news.ycombinator.com/item?id=50602551)

> Finally someone created the "Uber but for LLMs"
> — [nielsole on hackernews · 1 comments](https://news.ycombinator.com/item?id=50601029)

**Source threads**

- [hackernews](https://news.ycombinator.com/item?id=50599762) · 223 points · 55 comments

## Similar posts on daily.dev

- [GitHub - aattaran/deepclaude: Use Claude Code's autonomous agent loop with DeepSeek V4 Pro, OpenRouter, or any Anthropic-compatible backend. Same UX, 17x cheaper.](https://daily.dev/posts/github---aattaran-deepclaude-use-claude-code-s-autonomous-agent-loop-with-deepseek-v4-pro-openrout-0rcoomwtj) · Hacker News · 3 upvotes · 0 comments

---

Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#llm](https://daily.dev/tags/llm), [#typescript](https://daily.dev/tags/typescript), [#openrouter](https://daily.dev/tags/openrouter)

[View this post on daily.dev](https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"Show HN: I built a game where you are an LLM's subagent","url":"https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc"},"datePublished":"2026-11-22T16:00:24.387Z","dateModified":"2026-11-22T21:31:47.734Z","description":"Worker 755 is a browser game where an AI manager, powered by a real LLM like Claude, GPT, Gemini, DeepSeek, GLM or Kimi, treats you as its subagent. The...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/92ffaa0ef0d819c406866226966731d5?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/92ffaa0ef0d819c406866226966731d5?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,llm,typescript,openrouter","timeRequired":"PT5M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"Show HN: I built a game where you are an LLM's subagent"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/show-hn-i-built-a-game-where-you-are-an-llm-s-subagent-tulcrrmoc#faq","mainEntity":[{"@type":"Question","name":"How do I run the worker755 game locally with Docker?","acceptedAnswer":{"@type":"Answer","text":"Build the image with 'docker build -t worker-755 .' then run it with 'docker run -p 8755:8755 -v worker755-data:/data --env-file .env worker-755'. The container listens on 0.0.0.0:8755 and stores shared reviews in a persistent volume mounted at /data. Setting TRUST_PROXY=1 is recommended when running behind Cloudflare. Exploring agent-orchestrator demos like this is easier with technical write-ups daily.dev surfaces for self-hosters."}},{"@type":"Question","name":"What LLMs can act as the manager in an AI agent orchestration game?","acceptedAnswer":{"@type":"Answer","text":"The manager can be powered by Claude, GPT, Gemini, DeepSeek, GLM, or Kimi, configured via the MANAGER_MODELS environment variable using gateway naming like 'vendor/model'. Players can also choose their own model when bringing an OpenRouter or Vercel AI Gateway key, with a full shift on Claude Haiku costing a few cents. Comparing which LLMs fit an orchestrator role is simpler when daily.dev keeps you current on new model releases."}},{"@type":"Question","name":"What Node.js version is required to run a server that executes TypeScript directly without a build step?","acceptedAnswer":{"@type":"Answer","text":"Node 24 or newer is required, since the server runs TypeScript directly without a separate compile step. Setup involves 'npm run setup' to install dependencies, 'npm run build' for the web app, and 'npm start' to serve on 127.0.0.1:8755 by default. Tracking which Node versions unlock native TypeScript execution matters when planning your stack on daily.dev."}}]}
```

