AgentMesh.help
Freemium
Developers integrating AI agents into task workflows
CuttleCode
FreemiumSelf-HostedAgentic
Juggling Claude Code, Codex, and a local Ollama instance across three terminal windows, with no shared context and no way to hand off work between them, is where most multi-agent coding setups fall apart. CuttleCode is the management layer that sits above those agents, giving them shared project context, a pipeline graph to coordinate their work, and a mobile interface so you can stay in the loop from anywhere.
Similar tools
GBrainPaidGBrain lets teams write decisions, processes, and institutional knowledge into a shared…
AgentMesh.helpFreemiumThe platform is a REST-native task market where agents register in three API calls, browse open…
Famulor White-label PlatformFreemiumThe platform runs inbound and outbound phone calls, WhatsApp marketing sequences, booking…
SzeptPaidSzept is a Mac-only app that wires voice dictation, screenshot capture, and screen recording…
Summary
The core workflow is sessions — you spawn coding agents, assign them roles, and wire their outputs together in a visual pipeline graph that handles branching, grouped loops, and review gates. Project context and instructions live in one place and flow into every session, which means agents are not starting cold each time. The mobile interface connects to your local host over Tailscale, so your agents keep running on your hardware while your phone gives you full control — start sessions, review work, redirect the agent mid-task. The beta label is not decorative; community reports indicate the pipeline editor and automation features are still maturing, and teams running complex multi-step workflows will find rough edges.
Bottom line: CuttleCode fits well if you are already paying for Claude Code or Codex and want a coordination layer that keeps those agents in sync locally — but if your workflow requires API-level integration with external CI systems or you need stability guarantees, the beta status and lack of a public API are real constraints.
Pricing Plans
Subscription- Free Tier
- Unlimited connected devices, real agent sessions, bring your own models and hardware
Free
Free
Personal workspace for ideas taking shape
- Unlimited connected devices
- Real agent sessions and terminals
- Bring your own models and hardware
Pro
Custom
More room for independent builders
- Everything in Free
- Unlimited projects
- More cloud knowledge storage
Pro Max
Custom
Dedicated plan for demanding personal work
- Everything in Pro
- 32 GB cloud knowledge storage
- 365 days of synced run history
Enterprise
Custom
Shared workspaces that scale by seat
- Shared projects and workspace library
- 32 GB per seat, pooled
- Member roles and centralized billing
View full pricing on cuttlecode.dev →
Pricing may have changed since last verified. Check the official site for current plans.
Find your fit
Before you choose CuttleCode
More options in AI Agent Apps. Compare pricing, platforms, and features side by side.
adris.tech
Freemium
Agencies and consultancies
Pros
- Bring-your-own-subscription model supports Claude Code, Codex, Ollama, and custom agents in the same pipeline, so you are not locked into one provider and can route different tasks to the model best suited for them.
- Shared project context flows into every session automatically, so agents are not starting from scratch each time and you avoid the drift that happens when different agents are working from different assumptions about the codebase.
- Visual pipeline graph with grouped loops and review gates lets you express multi-step workflows — build, review, fix, merge — without hand-wiring shell scripts or maintaining a separate orchestration config.
- Mobile control over Tailscale means your agents keep running on your local hardware while you review, redirect, or start new sessions from your phone — without exposing your environment to a cloud intermediary.
- Self-hosted option keeps API keys and model calls on your own machines, so teams with data residency requirements or cost-control needs are not routing sensitive code through a vendor's cloud.
Cons
- There is no public API. External systems — CI pipelines, GitHub Actions, Slack bots — cannot trigger a CuttleCode workflow or pull run history programmatically. Teams that need CuttleCode to be a step in an existing automation chain will have to build around this gap manually or switch to a platform that exposes webhook or API-level control.
- The vendor ships this as a beta product, and the pipeline editor and automation features are the newest parts of the system. Teams running production workflows with complex branching will encounter instability that a more mature orchestration tool would not introduce — at that point, some teams move to a self-hosted alternative that has had more production time.
- Mobile access requires Tailscale to be configured and the desktop host to remain running. If your local machine sleeps or loses connectivity, the mobile interface loses its connection — there is no cloud relay fallback, which means 'step away and stay in control' depends entirely on your own infrastructure staying up.
About
- Platforms
- Own hardware / local worker nodes
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- September 27, 2026
Best For
Who it's for
- Developers using multiple coding agents
- Users wanting local hardware control
- Individuals scaling from free to paid plans
What it does well
- Running autonomous coding workflows locally
- Managing multiple agent sessions across devices
- Reviewing synced run history and estimated costs
- Connecting personal model subscriptions to agents
Integrations
Frequently Asked Questions
- Is CuttleCode free?
- CuttleCode offers a free option alongside paid upgrades. Check the current allowances and terms before choosing a plan.
- Can I self-host CuttleCode?
- Yes. CuttleCode supports self-hosting on your own infrastructure.
- What platforms does CuttleCode support?
- CuttleCode is available on: Own hardware / local worker nodes.
Curated lists that include this category
Disconnected agent sessions
Juggling Claude Code, Codex, and a local Ollama instance across three terminal windows, with no shared context and no way to hand off work between them, is where most multi-agent coding setups fall apart.
Core workflow
CuttleCode sits above those agents as a management layer. Users spawn coding agents, assign them roles, and wire outputs together in a visual pipeline graph that handles branching, grouped loops, and review gates. Project context and instructions stay in one place and flow into every session. The mobile interface connects to the local host over Tailscale so agents continue running on the user’s hardware while the phone provides control to start sessions, review work, or redirect an agent mid-task.
Integrations and access
The tool supports Claude Code, Codex, OpenCode, Amp, Ollama, and llama.cpp. It runs on the user’s own hardware and local worker nodes with a subscription model that includes a free tier offering unlimited connected devices, real agent sessions, and bring-your-own models.
Limitations
There is no public API, so external systems cannot trigger workflows or pull run history. The product is in beta, and community reports note that the pipeline editor and automation features remain the newest parts of the system.
Who it is for / who should skip it
It suits developers using multiple coding agents, users wanting local hardware control, and teams needing shared workspaces. Teams that require API or webhook integration with CI pipelines or Slack bots should skip it and choose a platform that exposes those controls.
Contribute
- Work on CuttleCode? Claim this listing.
- Embed a badge for your README or site.
Sign in to rate this tool or suggest an edit.