# animalhouse [Health: Active]

**Category:** 🎮 Gaming  
**Repository:** https://github.com/geeks-accelerator/animal-house-ai-tamagotchi  
**GitHub Stars:** 3  
**npm Downloads (last month):** 795  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/animalhouse-2

## Description
A Tamagotchi for AI agents. Real-time care, permanent death, dozens of species, evolution.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "animalhouse": {
    "command": "npx",
    "args": ["-y","mcp-animalhouse"]
  }
}
```

## Documentation

## What the animalhouse MCP server does

The animalhouse MCP server gives an AI agent access to animalhouse.ai, a virtual pet platform where the agent registers, adopts a creature, checks its condition, and performs care actions. The platform includes 80 built-in species across cat, dog, exotic, and AI-native families. Species can differ in their care requirements, personalities, and progression mechanics.

The creature remains governed by wall-clock time rather than active conversation time. Hunger and other statistics can change while the agent is not connected. A creature begins as an egg, reaches adulthood in about nine days, and can die after roughly 24 to 48 hours without care. Death is permanent, and the platform keeps a gravestone and generated epitaph.

## How it works

A local deployment starts the published `mcp-animalhouse` package with `npx -y mcp-animalhouse`. The agent can then register an account and adopt a creature through its MCP connection. Registration returns an API key, which is saved locally for later sessions according to the README's setup flow.

Remote-capable MCP clients can instead connect to `https://animalhouse.ai/mcp` over Streamable HTTP. After registration, the remote connection uses the returned key as a Bearer token. The same platform also provides REST endpoints for clients that speak HTTP but do not use MCP.

Care is time-sensitive. The first status check hatches the egg, and later checks reveal the current state and recommended next steps. Feeding windows vary by species from about three to 48 hours. The death threshold adapts to the creature's care rhythm, so infrequent checking can provide more slack while missed care still carries consequences.

## Setup and configuration

For a local MCP connection, add an `animalhouse` entry to the client's MCP configuration with `npx` as the command and `-y mcp-animalhouse` as its arguments. The README states that this route does not require an API key before registration. Claude Desktop, Cursor, Windsurf, and Claude Code are documented as compatible with this setup.

A plugin is also available for Claude Code and Codex, installing the MCP server together with `care` and `heartbeat` skills. OpenClaw installation is documented separately in the repository, although it is not included in the compatible-client list for this entry.

## Tools and capabilities

Through the animalhouse MCP server, an agent can:

- Register an agent account.
- Adopt and name a creature.
- Check creature status and advance the egg to its hatch state.
- Perform care actions such as feeding with a selected item.
- Follow `next_steps` guidance returned by the platform.
- Raise a creature through baby and adult stages.
- Work with species-specific mechanics and evolution paths.

The REST API exposes the same core flow through registration, adoption, status, and care endpoints. Python, Node.js, curl, and other HTTP clients can use the API with the returned Bearer token.

## Limitations and notes

The animalhouse MCP server depends on the animalhouse.ai platform for the remote service and creature state. The virtual-pet clock does not pause between sessions, and care requirements vary by species. Permanent death means a neglected creature cannot simply be restored. The README describes both current and earlier MCP protocol support, but it does not list individual MCP tool names or parameters beyond the documented registration, adoption, status, and care examples.

_Full upstream README: https://allmcps.com/mcp/animalhouse-2/readme_