OpenPets visualizes AI agent progress with MCP companions

OpenPets is the Tamagotchi moment for the agentic era, solving the "black box" problem of autonomous agents through low-friction visual metaphors.

• –MCP integration allows first-class support for Claude Code and OpenCode, with tools that let agents "speak" or trigger specific emotional reactions.
• –Privacy-first design filters sensitive code and paths from speech bubbles, ensuring the companion stays decorative rather than a data leak vector.
• –The three-layer integration (MCP tools, instructions, and hooks) provides a blueprint for how future developer tools should handle ambient state monitoring.
• –While purely aesthetic, the psychological impact of seeing a pet "celebrate" a successful test run significantly improves the developer experience in high-latency agent loops.
• –Support for custom pet packs and a Swift library for native macOS embedding points toward a broader ecosystem for agent-aware UI components.