A Field Guide to Digital Companionship Infrastructure

What it does

This is a curated directory of 157 open-source projects built for long-term AI companion relationships. It covers chat clients, memory systems, proactive messaging runtimes, voice and embodiment tools, virtual phones, and shared activity platforms. Each entry is tagged by language, platform, and a readiness level—ready, adapt, infra, or verify—with descriptions pulled from project READMEs rather than repo names alone.

The interesting bit

The taxonomy treats companionship as a systems engineering problem. Categories like “Background Heartbeats & Proactive Messaging” and “Memory, Identity & Emotion State” separate the mechanics of persistence from the chat interface itself, making it easier to find the specific glue you need.

Key highlights

• 157 entries tagged by language, platform, and readiness (ready, adapt, infra, verify).
• Covers niche categories like virtual phones, desktop pets, proactive heartbeats, and long-term memory engines.
• Descriptions include real caveats: some projects are early WIP with simulated replies, others are deeply tied to their author’s personal canon.
• Includes native mobile clients, self-hosted LAN hubs, browser PWAs, and CLI coding-agent companions.
• Explicitly flags entries with thin public documentation so you know when to re-check before relying on them.

Caveats

• This is a catalog, not a framework; you browse it, not install it.
• A few verify-tagged projects are essentially mockups—one virtual phone admits its current replies are simulated with no LLM wired in yet.
• Several entries are heavily personalized to their creators and flagged as better architecture references than reusable apps.

Verdict

Bookmark this if you’re building—or just morbidly curious about—the plumbing behind persistent AI companions. Look elsewhere if you want a single turnkey app; this is a map, not a destination.