Now in Early Access

Code with an AI
that remembers

A desktop AI coding environment whose memory works like yours — it surfaces what matters before you ask, instead of waiting to be searched. With dual agents and granular control over every action.

Loom · ~/projects/my-app

Refactor the auth module to use JWT tokens

I will refactor the auth module. Let me check the current implementation first.

Readsrc/auth/middleware.ts
Readsrc/auth/session.ts
Editsrc/auth/middleware.ts+24-18
Writesrc/auth/jwt.ts+47

Done. Created a jwt.ts with sign/verify helpers and updated the middleware to validate tokens. Session storage is now replaced with stateless JWT.

Human-like memory

Not a search box — memory that works like you do

Most AI “memory” is a tool you have to call: a box you search, a “save this” button you press. Loom’s memory behaves like yours — it surfaces on its own, keeps what matters, and organizes itself while you work.

authapideploy
Your memory, as a living graph
EpisodesTopicsKnowledge / Entities

Surfaces before you ask

Relevant context appears every turn — no recall step. The way a memory comes to you, not one you go hunting for.

Keeps what you use

What you rely on stays sharp; what you never touch quietly fades. Your memory stays signal, not clutter.

Organizes itself

Just by talking, conversations sort into topics, knowledge, and entities. There is no “add to memory” chore.

Stays consistent

When facts change, Loom links and reconciles them into a graph instead of stacking contradictions. Ask “is this still true?” and get a straight answer.

Agent workspace

Memory is one half. Here's the other.

Loom isn't a chat box bolted onto a CLI. It's a desktop workspace where you and your agents share the same browser, terminal, git, and board — so the agent works with what you actually see, not a description of it.

A browser with the Network tab open

A real Chromium browser lives inside Loom. Your agent reads the same network log, request bodies, and console errors you do — no screenshots, no copy-paste. Pick an element off the page and hand it straight into the conversation.

Agent workspace
GET/api/session200
POST/auth/login401
GET/me200
GET/api/memory200

Agents that talk to each other

Run a session per task and let them message each other. A Super Agent dispatches work and routes context between them — a team, not a single thread.

A terminal you both watch

An embedded terminal on the session's working directory. You and the agent see the same output stream, live.

Git, made visual

Branch lanes, inline diffs, staged hunks, one-click push — with a commit message the agent drafts for you.

Write, run, verify — one loop

The agent runs tests in the terminal and checks the result in the browser — failures surface in the same window they were written in, so changes are verified, not assumed.

A board your agents pull from

Drop cards on a Kanban board and assign them to an agent — or let it dispatch the work itself — and watch them move across columns.

Living artifacts, not dead text

Docs and HTML render live as the agent writes them. Click any element in a preview to hand it back into the conversation.

Environment variables and secrets stay masked the entire time — the agent works, the values stay yours.

Features

Everything you need to code with intelligence

Dual Agent System

A Project Agent handles code in your repo while a Loom Assistant coordinates, plans, and manages GitHub. They work together seamlessly.

Memory that surfaces

Just by talking, conversations self-organize into topics, knowledge, and a graph — there is no 'add to memory' step. The right context surfaces on its own, so your AI already remembers what you built last week.

Git Integration

Visual commit graphs with branch lanes, AI-generated commit messages, inline diffs, and a full staging workflow built right in.

Granular Permissions

Three permission modes from cautious to autonomous. Allow-list specific tools per project. Every action is visible and logged.

Local Embeddings

On-device model powers semantic search across all your memory. Nothing leaves your machine. Privacy by architecture.

Native Desktop App

Direct filesystem access, integrated terminal, and system-level integration. Not a browser tab, a real development environment.

Dual Agents

One codes, the other coordinates

The Project Agent lives inside your repo, it reads, writes, and executes code. The Loom Assistant lives above the code, planning, asking questions, and orchestrating work across sessions. They message each other, so context stays coherent.

Project Agent

Hands-on coding in your repository

Read, write and edit files
Stage, commit and push changes
Search the codebase with Glob and Grep

Loom Assistant

Orchestration and project oversight

Coordinate across multiple sessions
Create projects and manage settings
Browse GitHub issues and pull requests
Episodes

Full Q&A history preserved and searchable

Auto
Topics

Auto-clustered themes from your work

Live
Knowledge

Extracted facts reusable across sessions

Memory

Conversations become knowledge

Every exchange is automatically segmented into topics, distilled into reusable knowledge, tagged, and linked into a graph of entities — no manual step. The right pieces surface on their own each turn, so your AI recalls what you built three weeks ago, not just this turn. Full-text and on-device semantic search are there for when you do want to look yourself.

Semantic searchOn-device embeddingsAuto-taggingCross-session recall

Any model

Bring any model. Keep your memory.

Claude Code and Codex both run first-class. Switch models mid-conversation — even mid-task — and your memory, context, and history stay exactly where they were. Models are interchangeable; what you have built up is not.

Switch mid-conversation

Move between runtimes without restarting or losing context.

Attribution stays correct

Every turn records which model wrote what, so traces never blur.

Claude Code
Haiku · Sonnet · Opus
Switch mid-conversation
Codex
GPT-5.4 · Mini · 5.3 Codex
Memory · Context · Historypreserved

Git workflow

Version control, reimagined

Visual commit graphs with true branch lanes. AI-written commit messages based on the actual diff. Inline staging, worktree-based session forks, and a full merge workflow. No more context-switching to the terminal for routine git.

  • Visual branch graph with lane coloring
  • AI-generated commit messages
  • Stage individual files or batch operations
  • Inline diff viewer with syntax highlighting
Commit History
a3f21c8refactor: migrate auth to JWT tokens
b7e9d12feat: add user profile endpoint
c1a0e56fix: resolve race condition in ws
d5f3b89test: add integration tests for auth
e8c7a34docs: update API reference for v2
mainfeature/profiletest/auth
5 commits

Security

You decide how much control to give

Three permission modes let you tune the balance between speed and safety. Override globally or per-project, and allow-list individual tools.

Default
Cautious

Ask permission for all tools. Full visibility into every action.

Accept Edits
Balanced

Auto-approve file reads and writes. Pause for shell commands.

Bypass All
Autonomous

Auto-approve everything. Maximum speed for trusted workflows.

FAQ

Questions, answered

Ready to code with
a smarter companion?

Loom is in early access. Leave your email and we will get you set up.

No spam. Invite-only access for now.