Central Brain for Agent Builders

One memory every coding agent reads and saves to

Ditto gives Claude Code, Cursor, Codex and your other agents one shared memory of your project. Every session starts from what was decided, and what an agent learns is saved for the next one.

Works with
  • Claude
  • ChatGPT
  • Cursor
  • Codex
  • Gemini
  • OpenClaw
  • Hermes
  • Any MCP client

What it looks like when your coding agents don't share a memory

  • Contradictory logic

    Agents give different answers about the same code because each works from its own notes.

    Rework plus regression risk
  • Amnesia on restart

    Restart a session and the agent has lost what it knew an hour ago.

    Files re-read and decisions re-made
  • Repeated mistakes

    A known gotcha gets hit again because the fix was never saved where agents look.

    Debugging time spent twice
  • Same context re‑explained every session

    You paste the same background into every new session before the real work starts.

    Tokens and minutes wasted up front
  • One agent undoes another's work

    One agent changes what another built because neither knows what the other did.

    Double spend and broken builds
  • Handoffs lose critical detail

    When work moves between agents or people, the details that matter drop out.

    Wrong output shipped to clients

What changes when every agent shares one central brain

Every agent you run reads from and saves to the same memory of your project.

Every agent works from the same decisions

Claude Code, Cursor and Codex read one memory, so they give the same answer about the same code.

Restart and pick up where you stopped

A new or restarted session opens with the decisions and progress already saved.

Fixes and gotchas are saved for next time

When an agent solves a tricky bug, the fix is saved where every agent looks first.

Nobody pastes a brief again

Every session starts with the project's context, so the first prompt is the real task.

Agents see what the others changed

Before an agent edits code, it reads what the other agents did and why.

Handoffs carry the full context

What an agent worked out is saved with the task, so the next agent or person starts from it.

How it works

Three steps, then every session starts with your project.

1

Connect your coding agents

Connect Claude Code, Cursor or Codex over MCP, or OpenClaw and Hermes through the CLI.

2

Ditto builds one project memory

It saves decisions, reasons and progress from your sessions, PRs and threads, and keeps them current.

3

Every agent starts from it

Each agent searches the memory before it starts and saves what it learned when it finishes.

Every AI and coding agent you use starts with the same memory

Connect Ditto once and stop re-explaining your work in each new session. Each tool searches the same memory and can save to it.

Any other MCP clientOther MCP clients, such as ChatGPT, can search and save the same memory.

Works with the coding agents and apps you already use. Over 1,400 integrations.

See all app integrations

Privacy and security for your code

Your code and your project memory stay yours, and every agent works inside the limits you set.

Your code and memory

  • Not used to train modelsDitto doesn't train on your code or your memory.
  • Separate memory per projectKeep each project or client in its own memory, so one never leaks into another.
  • You choose what each session usesPick which memories a message can draw on.
  • Memory that can leave with youExport your whole memory as a signed Ditto Passport, or delete any part of it.

Control over every agent

  • Read-only to startDitto reads only the repos and apps you connect. You switch on writing app by app.
  • Agent jobs wait for approvalDitto prepares the job, and it runs only after you approve it.
  • See what every agent didThe activity log shows each memory an agent read and saved, and what it changed.
  • Your provider's terms stay in placeWhen Claude Code or another tool reads your memory, that provider's policies apply to what it reads.

Ditto's memory harness and integrations are open source.How Ditto handles your data

Start free, then pick a plan

Your plan covers capacity and features. Credits pay for variable usage across AI, voice, tool actions and storage beyond your included allowance.

New plan preview. Rollout is in progress; check the offers available in your account before purchasing.

Free

$0 No card needed

Explore Ditto with no card needed.

1B starting credits; earn up to 4B1 GB included storage5 saved workflows · 1 active

Plan includes:

  • 1 space
  • Chat in Ditto
  • Auto model selection
Optional: Flex pay as you go

Add a $5 initial deposit for more usage, with no subscription. Net credits follow the disclosed payment-fee reserve.

Paid chat connections, Ditto Review and remote control unlock after the deposit settles.
Try Ditto free
Everyday work

Plus

$19/month Monthly subscription

More room for your everyday work.

$10 usage each month · 10B credits50 GB included storage10 saved workflows · 5 active

Plan includes:

  • 3 spaces
  • Paid chat connections
  • Ditto Review
  • Auto model selection
Try Ditto free

Pro

$97/month Monthly subscription

Shared capacity and credits for you and your team.

$60 usage each month · 60B credits1,000 GB included storage100 saved workflows · 25 active

Plan includes:

  • 10 spaces
  • Paid chat connections
  • Ditto Review
  • 1 org · 4 included licensed seats
  • Choose models manually
  • Custom domains
Try Ditto free

Enterprise

A custom Ditto build for your business: our team writes the spec, builds it and supports it.

Custom setup and retainer Get in touch

Questions about coding agent memory

What is coding agent memory?

Coding agent memory lets an agent keep what it learns about your project and use it in later sessions. Ditto keeps one memory that every agent you connect can search and save to.

Which coding agents work with Ditto?

Claude Code, Cursor and Codex connect over MCP. OpenClaw and Hermes connect through the Ditto CLI. Any other MCP client can connect too.

How is this different from CLAUDE.md or a rules file?

A rules file holds what you wrote down and stays the same until you edit it. Ditto saves decisions and progress as agents work and keeps them current, and every agent reads the same memory.

Can my agents read each other's memory?

Yes. What one agent saves, the others you connect can read, so Cursor can use what Claude Code learned.

Does it work if I switch models?

Yes. The memory lives in Ditto, not in one model or tool, so a new model starts from the same project memory.

Does Claude Code have its own memory?

Yes. Claude Code keeps memory for its own sessions. Ditto adds one memory that Claude Code shares with your other agents and tools.

Is my code used to train AI models?

No. Ditto doesn't train on your code or your memory. When an AI tool reads your Ditto memory, that provider's policies apply to what it reads.

What does Ditto cost?

The new plans are Free, Flex with a $5 initial prepaid deposit and no monthly base fee, Plus at $19/month, and Pro at $97/month. Usage is prepaid with credits. See the pricing page for plan details and rollout availability.

Get 20% off your first month of Ditto, and get more out of it every week.