Deep dive

The rest of the command center.

Everything beyond the front page: how context is composed, how skills improve, what starts work, where it runs, and what stays on the record.

Context is the foundation

A project is bigger
than a repository.

The frontend, the API, the shared library. Your agents need to understand how they fit together. CuttleCode gives that understanding a place to live.

See the context walkthrough
Workspace
Atlas StudioPeople, credentials, and workspace defaults
Project
Atlas CommerceOne purpose. Shared instructions.
  • storefrontReact
  • payments-apiNode
  • design-systemShared UI
  • Project rules
  • Skills
  • Agents
  • Knowledge

An example structure · inspect the real one in the tour

Context with a home.

Keep project instructions, repo references, skills, and agents together. Reuse what matters across sessions without rebuilding the briefing every time.

See what goes in.

The context harness shows the composed instructions, context layers, token estimates, and authoring findings. Inspect the starting point before asking an agent to work.

The right scope for the work.

Start at the project to reason across repositories, or work in a specific repo. Give specialized agents the skills they need for their part of the job.

Inside your workspace

Meet your
command center.

Keep sessions, agent loops, and their progress together. Open a capture for a closer look.

Forge

Better skills. Tested, not guessed.

Turn the instructions your agents use into versions you can compare. Forge drafts alternatives, evaluates them against reserved cases, and brings eligible improvements back for your review.

Skills, agent prompts, and Ralph strategies can all enter Forge. Scheduled evolution proposes improvements; automatic promotion is a separate opt-in for artifacts you mark low-risk.

Illustrated skill review
01 / 06
Keep a baseline

The current version—the champion—keeps serving your sessions while Forge works on alternatives.

Work can start anywhere

Your workflow.
Connected end to end.

Start with the tools your team already uses. Bring the request into a project with the context and permissions to act on it.

  1. 01Linear story
  2. 02Cuttle intent
  3. 03Autoplay
  4. 04PR for review
Or meet your team in Slack and Discord
01

A Linear story becomes the next intent.

Bind a project to Linear, choose the assignees to follow, and adopt matching stories as intents. Autoplay carries one from research toward a PR, with progress and comments tied to the original story.

  • Assignee filters
  • Autoplay
  • Two-way updates
02

On a schedule. Or on a signal.

Run a recurring review on a cron schedule, or start work from an incoming webhook. Either can trigger a session or a pipeline. Webhooks need a rotatable bearer secret — the URL alone is not enough.

  • Cron & intervals
  • Webhook secrets
  • Sessions & pipelines
03

Your internal app, with project context.

Expose a project’s HTTP service to build an internal agent chat. Map guest and signed-in users to gated actions, load the context each role needs, and enforce tool permissions at the agent boundary. You configure it, with your own authentication and authorization.

  • Project HTTP services
  • Identity-aware access
  • Tool restrictions

Multiple workspaces

Separate workspaces
for separate work.

Keep each client, company, or personal project in its own workspace, with separate credentials, members, and settings.

Client workAtlas Studio

Projects
Team projects
Credentials
Workspace Git credentials
Defaults
Shared skills & defaults

Your next ideaAfter Hours

Projects
Personal projects
Credentials
Separate integrations
Defaults
Your own settings

The details add up

Made for the way
you actually work.

Quiet features you end up using every day.

Workspace boundaries

Keep engagements separate.

Projects, members, Git credentials and model defaults live per workspace.

With a cloud account

Link machines and team.

Unlimited devices on every cloud plan. Invite and manage members.

Your hardware

Stay in control of execution.

Choose what syncs per machine. Transcripts stay on the instance.

On your phone
Projects, intents and sessions, on mobile.
The Cuttle CLI
Ask a session to find past work, create an automation, or hand off to another agent. Script it with API keys, too.
Caveman Mode
Terse replies for the sessions you launch.
Ponytail
Simpler code first. Lite, Full, or Ultra.
Trusted remote hosts
Work over SSH, near the repos and compute.
Notifications
Discord, ntfy, browser push, or email.
Any Git host
GitHub, GitLab, Bitbucket, Azure DevOps, or another Git host.
Even Realities G2
Cuttle within reach through the G2 companion.

Knowledge & platform

Your machines. A shared workspace.

Knowledge
Build on what your team already knows.
LLM Wiki
A living knowledge base auto-linked to projects and sessions.
Ask the Codebase
Local-embedding RAG with file:line citations verified at answer time.src/pty/manager.ts:214✓ verified
Platform
Pool the hardware. Route every job.
Worker Nodes
Pool your GPUs. Dial-home nodes serve local models; CuttleCode routes each job to the best fit.
Media
Images and video on your own stable-diffusion.cpp host. Bring your models — FLUX, ERNIE, Wan2.2, whatever you run.
Local inference
Flip the router from cloud to local models. Offline mode built in.cloud → localoffline mode
Computer Use
Isolated desktop control on Linux, Windows, macOS, or Docker — never your operator console.

Surfaces

Take the command center with you.

Manage work from desktop or phone. Start sessions, follow progress, and keep your agent choices and account preferences close at hand.

Your host. Your connection.

Built-in Tailscale support takes your home-hosted instance beyond your home network. Set up Tailscale on your host and phone, or host CuttleCode online on your own VPS. You run the instance and manage its connection.

Reaches
  • Slack
  • Discord
  • MCP
  • Claude Code
  • Codex
  • Gemini CLI
  • OpenCode
  • Amp
  • Ollama
  • llama.cpp
  • Chrome extension
  • Cloud workspaces
  • CLI
  • tray
  • Local Agent PWA
  • Remote Hosts
  • Pipelines
  • Ralph loops
  • Triggers
  • Worktrees
  • Wiki
  • Webhooks
  • REST API
G2 glasses
See which sessions need you, read their replies, and dictate a response from Even Realities G2 glasses. Swipe to browse. Tap to open.
Even Realities G2Illustrated glasses view
Your sessions, within sight. Open a conversation to read or dictate a reply.

browser

Work in the browser you already use.

Pair the Chrome extension, explicitly share a tab, and let a human-started session work with the page in your signed-in browser. Stop sharing when you’re done.

Requires extension setup and browser support enabled in Cuttle. Only shared tabs and granted sites are reachable; scheduled automation cannot drive this browser lane.

workspaces

A home for your projects, people, and devices.

Keep personal and shared workspaces separate. Link your machines, invite collaborators to shared workspaces, and manage projects and devices from your account. Cloud sync can be controlled per workspace on each machine.

Shared workspaces and cloud limits depend on your plan. Cloud run history carries status, verdicts, and cost summaries; session transcripts and terminal output stay on the instance.

Governance

Set the boundaries. Keep the record.

Control who can act, give each session the right context, and keep the evidence close to the work.

See what CuttleCode stores
  1. Before access

    Decide who can act

    Set roles and permissions for your workspace. Keep credentials in encrypted storage.

  2. At session launch

    Bring the right context

    Project instructions and scoped skills follow the work across supported agent harnesses.

  3. During the work

    Keep changes separate

    Use isolated worktrees so concurrent changes stay separate until you review them.

  4. After each run

    Follow the evidence

    Inspect audit events, decision records, and token usage alongside the work they describe.

Download the beta

Bring the idea. Take the helm.

Try the CuttleCode beta. Your agents, your hardware, your final say.