install

Additional install commands

macOS

Homebrewverified · 100%

brew install letta-code

local Homebrew formula metadata

Windows

Windows Package Managerverified · 92%

winget install --id Letta.LettaCode -e

Windows Package Manager source index · Letta.LettaCode · source: cdn.winget.microsoft.com

overview

Package summary

Memory-first coding agent

Project links

Commands and aliases

letta

history

Project history and usage

Letta Code is a memory-first coding-agent CLI and harness from Letta. Its project story belongs to the post-2024 wave of terminal coding agents, with the differentiator that agent identity, memory, skills, and conversation history are treated as durable state rather than per-session context.

Project history

The `letta-ai/letta-code` repository was created on October 25, 2025. Letta's product documentation presents the tool as the open-source Letta Agent harness for coding agents, desktop agents, personal agents, and more autonomous workflows.

Letta's launch material framed Letta Code as a response to stateless coding-agent sessions: instead of restarting from scratch, the same agent can learn project conventions through commands such as `/init`, `/remember`, and skill creation.

Adoption history

Letta Code entered package-manager culture quickly for a young AI developer tool: this batch records Homebrew and winget package identifiers, while the official quickstart documents global npm installation with `npm install -g @letta-ai/letta-code` and execution through the `letta` command.

How it is used

The intended workflow is to start `letta` inside a project, connect model providers or coding plans, and let the agent inspect the codebase, retain memories, and resume future conversations. The docs also describe headless prompts, remote environments, subagents, skills, and memory-management commands.

Why package nerds care

For package nerds, Letta Code is notable less as a stable old Unix utility and more as a fast-moving example of AI tools becoming ordinary CLI packages. Its package metadata, npm install path, and Homebrew/winget entries show the distribution path for AI coding agents moving into the same channels as shells, editors, and build tools.

Timeline

2025: The `letta-ai/letta-code` repository was created on October 25, 2025.
2026: Letta documentation described Letta Agent/Letta Code as an open-source, memory-first agent harness with CLI, desktop, and cloud setup paths.
2026: The Letta Code changelog documented active 0.x releases, including 0.27.11.

Related projects

Letta Code is part of the broader Letta/MemGPT lineage around long-term memory for agents. It is also positioned beside terminal coding agents such as Claude Code, Gemini CLI, and Codex CLI, while using npm, Homebrew, and winget as ordinary developer-tool distribution channels.

Sources

security posture

No protected-tool coverage found yet

No matching local secret-handling manifest was found for letta-code. Nucleus package metadata is still published here so future coverage has a stable package URL.

Install behavior

No Homebrew post-install hook is recorded in formula metadata.
Homebrew bottle metadata is available for 6 platform targets.
Installs with 5 runtime dependencies.
Build metadata lists 1 build dependencies.

Recommended review

Before unattended agent use, check whether the tool reads plaintext credentials, writes remote state, publishes artifacts, or shells out to plugins.

local files

Configuration and credential file locations

These source-backed paths show where this package keeps local settings or durable credentials. Automic Vault can use them as review targets for secret scanning, migration, and command approval.

Configuration files

Config paths the tool may read or write during local use.

Unix: ~/.letta/channels/<channel>/channel.json~/.letta/channels/<channel>/routing.yaml~/.letta/channels/<channel>/pairing.yaml

Credential files

Credential-bearing paths to review before unattended agent runs.

Unix: ~/.letta/channels/<channel>/accounts.json

executables

Installed executables

Command	Kind	Exposure	Note
`letta`	cli	global executable

freshness

Version and freshness

These signals separate page generation age, package-manager activity, and upstream release comparison. Version lag is warned only when an evidence URL and comparable versions are present.

page generated2026-07-08

manager version0.27.27

manager updated2026-07-08

local dataok

upstreamnot checked

latest detectednot detected

https://docs.letta.com/letta-code

infoRelease/tag comparison is only available for GitHub repositories.https://docs.letta.com/letta-codenone confidence

install metadata

Package metadata

Package key	brew:letta-code
Version	0.27.27
Package manager	Homebrew
Package manager page	https://formulae.brew.sh/formula/letta-code
Homepage	https://docs.letta.com/letta-code
Repository	https://github.com/letta-ai/letta-code
Upstream docs	https://docs.letta.com/letta-code
License	Apache-2.0
Source archive	https://registry.npmjs.org/@letta-ai/letta-code/-/letta-code-0.27.27.tgz
Last updated	2026-07-08T03:44:27Z
Pulse	updated
Dependencies	gettext, glib, node, ripgrep, vips
Build dependencies	pkgconf
Bottle	available (on arm64_linux, arm64_sequoia, arm64_sonoma, arm64_tahoe, sonoma, x86_64_linux)
Homebrew post-install	not defined
Service	none declared

registry facts

Source database details

Source Database	Homebrew formula API
Tap	homebrew/core
Full Name	letta-code
Version Scheme	0
Revision	0
Bottle Stable Root URL	https://ghcr.io/v2/homebrew/core
Deprecated	no
Disabled	no
Keg Only	no
URL Keys	stable

source database matches

Other package-manager records

Matches are pulled from external package-manager indexes and kept separate from local Automic Vault package links.

winget95%

Letta.LettaCode

winget install --id Letta.LettaCode -e

normalized package name match
Matched by: Letta Code

Windows Package Manager source index · cdn.winget.microsoft.com · Windows Package Manager source index: Letta.LettaCode from https://cdn.winget.microsoft.com/cache/source.msix

package graph

Links come from deterministic package relationships, av.db category and tag curation, ecosystem matches, and package hub membership.

source trail

Generated from repository data

This page is generated by av-web from the private package SQLite artifact built by scripts/generate-pkg-sqlite.py.

Combined YAML source

View the package source record on GitHub.

combined/letta-code.yml

Used sources

Geiger risk classifier
Nucleus package database
av.db category and tag curation
cross-ecosystem install command graph
curated configuration and credential file locations
curated package history
external package-manager database matches
package relationship graph
package version freshness
package-page enrichment

Install letta-code with Homebrew, winget