#413 · Primary category: AI Agents & Automation

letta-code

agent-memory ai claude codex continual-learning letta memgpt stateful-agents

Stateful agents that are like people, with memory, identity, and the ability to learn and adapt

Project last updated:08/29/26

GitHub Stars

3.2K

Forks

380

Contributors

49

License

Apache-2.0

Why we included this project

Most AI assistants wipe the slate clean when a session ends, but Letta Code agents are built to carry context forward. This open-source harness gives each agent a memory system it rewrites on its own, including its own system prompt and skills, so it accumulates experience and gets better across long, multi-session interactions. You can reach it through a local CLI, a desktop app, a browser, or chat platforms like Slack, Telegram, and Discord, which makes it a reasonable fit for always-on assistants that work proactively instead of only answering when called. It also includes scheduling, hooks, subagents, and permission controls, so teams can put durable agents into workflows that need context and identity preserved over weeks rather than rebuilt from scratch each time.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category