Index / LettaBot · updated Oct 2, 2026
LettaBot
letta-ai/lettabot · abandoned · rank 32 of 43 by stars
A multi-channel personal AI assistant with unified persistent memory across Telegram, Slack, Discord, WhatsApp, and Signal, built on the Letta platform. Now officially deprecated in favor of Letta Code's Channels and Remote Environments.
Facts
Repository
Runtime
Posture
Community
Security breakdown
Composite 55 / 100 · how these are scored
higher is safer
higher is safer
higher is safer
higher is safer
higher is riskier
Model access
Read from the repository, not written by a model · 60 files examined
2 providers · custom endpoint
Pinned models
claude-sonnet-4-20250514 released 2025-05-14 — 17 months old (dated in the model id)
Pin last edited 8 months ago
Evidence
Decision
Why choose LettaBot over OpenClaw?
Why choose this
- Unified persistent memory across five messaging platforms
- Backed by the Letta stateful agent platform with managed memory
- Voice transcription and Bluesky feed ingestion built in
Tradeoffs
- Officially deprecated and unmaintained
- Requires a Letta API key or Letta server dependency
- Local shell/file tool execution with limited sandboxing
Best fit
- Existing LettaBot deployments that still work
- Users wanting unified memory across many chat platforms
- Letta platform users exploring legacy integrations
Avoid if
- Starting any new project (officially deprecated)
- You need long-term maintenance and security patches
- You want a self-contained local-first assistant without a Letta backend
Evidence is strong that the project is deprecated (explicit README warning and archive notice), but community sentiment data is thin with zero Reddit matches and only indirect web mentions.
AI layer reviewed Sep 21, 2026 · how this is written
Star activity
326 stars today
Overview
LettaBot is a personal AI assistant built on the Letta platform that connects a single stateful agent to Telegram, Slack, Discord, WhatsApp, and Signal, with read-only Bluesky Jetstream feed ingestion. Its standout feature is unified persistent memory: one agent remembers conversations across all channels and across sessions spanning days or months, powered by the Letta Code SDK. It also supports local tool execution (file reads, code search, shell commands), voice message transcription via OpenAI Whisper, heartbeat check-ins, and agent-created scheduled reminders.
Architecturally, it is a Node.js/TypeScript application that requires either a Letta cloud API key or a self-hosted Letta Docker server, making it dependent on the Letta backend rather than fully local-first. Configuration is YAML-driven, and a Railway one-click deploy template lowers the barrier for always-on hosting.
Critically, LettaBot is officially deprecated. The maintainers have archived the project and direct all new development to Letta Code's Channels (for messaging integrations) and Remote Environments (for always-on agents via outbound WebSocket). Existing deployments continue to function, and migration is described as taking roughly 20 minutes, but anyone evaluating OpenClaw alternatives should treat this as a legacy reference implementation rather than a going concern.