Charles Packer 1c663dd0c8 Update README.md
2025-12-14 15:24:46 -08:00
2025-10-24 21:19:24 -07:00
2025-11-26 19:28:26 -08:00
2025-12-14 14:27:31 -08:00
2025-10-24 21:19:24 -07:00
2025-10-24 21:19:24 -07:00
2025-11-26 11:39:55 -08:00
2025-10-24 21:19:24 -07:00
2025-12-12 17:16:09 -08:00
2025-10-24 21:19:24 -07:00
2025-12-14 15:24:15 -08:00
2025-10-24 21:19:24 -07:00
2025-12-12 17:16:09 -08:00
2025-12-14 15:24:46 -08:00

Letta Code (Research Preview)

Letta Code is a self-improving, stateful coding agent that can learn from experience and improve with use. You can use Letta Code as a general purpose CLI harness to connect any Letta agent (even non-coding agents!) to your local dev environment.

Letta Code is open source and model agnostic - supporting Claude Sonnet/Opus, GPT-5, Gemini 3 Pro, GLM-4.6, and more.

Read more about how to use Letta Code on the official docs page.

letta-code

Important

Letta Code is a research preview in active development, and may have bugs or unexpected issues. To learn more about the roadmap and chat with the dev team, visit our Discord. Contributions welcome, join the fun.

Get started

Requirements:

Install the package via npm:

npm install -g @letta-ai/letta-code

Then run letta to start Letta Code in interactive mode (see various command-line options on the docs):

letta

If you haven't used Letta Code before, you'll need to use OAuth to login, or set a LETTA_API_KEY in your environment variables.

To connect to an existing agent, use the --agent flag:

letta --agent [existing-agent-id]

You can also run Letta Code in headless mode, making it easy to integrate into scripts (see the docs for more):

letta -p "Look around this repo and write a README.md documenting it at the root level"

Memory and Skill Learning

All agents in Letta are stateful: they maintain context forever and can self-edit their own memory blocks.

If youre using Letta Code for the first time, you will likely want to run the /init command to initialize the agents memory system:

> /init

Over time, the agent will update its memory as it learns. To actively guide your agents memory, you can use the /remember command:

> /remember [optional instructions on what to remember]

Skills are reusable modules that teach your agent new capabilities. Theyre automatically discovered from your projects .skills directory and loaded into the agents memory at session start. The easiest way to create a skill is using the interactive skill creation mode:

> /skill

Read the docs to learn more about skills and skill learning.


Made with 💜 in San Francisco

Description
letta-code - primary development repo
Readme 37 MiB
Languages
TypeScript 98.2%
Python 1%
Shell 0.5%
JavaScript 0.2%
MDX 0.1%