For engineers with more ideas than evenings
Stop babysitting chats. Steer a team of agents from one board.
You can't keep eight Claude Code tabs alive and still think. ampm runs every idea through planning, building, testing and review in parallel, on the subscription you already pay for, and only interrupts you for the calls that are actually yours: the architecture, the trade-offs, the merge.
A kanban board with four lanes — Ideating, Planning, Building and Shipping — carrying tasks from a fresh idea through to a pull request waiting for review.
Async, not synchronous
You were never meant to be the agent's scheduler
A chat agent is fast right up until it needs you, and then it waits — mid-thought, in a tab you forgot about. Multiply that by every idea you want to build and you're not building, you're babysitting.
- Parallel work
- Babysitting chats:As many sessions as you can hold in your head at once.
- ampm:As many as your subscription allows, each in its own sandbox and worktree.
- Your attention
- Babysitting chats:Every tab needs you the moment it gets stuck.
- ampm:One inbox. Answer when you surface; nothing is lost while you're away.
- Context
- Babysitting chats:Re-explain the project at the top of every session.
- ampm:Plans, decisions and your repo's rules persist with every task.
- Steering
- Babysitting chats:Paste corrections and hope it picks them up.
- ampm:Message a running agent, stop it mid-turn, or send work back with a reason.
- Stepping away
- Babysitting chats:Close the laptop and the work stops.
- ampm:Work continues overnight, and you're notified with no tab open.
You keep the wheel
You make the calls that matter. Agents do the typing.
When a planner hits a real trade-off — last-write-wins or a CRDT, a queue or a cron — it stops and asks, with its recommendation and its reasoning attached. The technical direction stays yours; the hours of implementation don't.
Nothing merges unless you want it to. Every ask carries the plan, the diff and the full transcript, so you decide in a minute what used to take a session.
- questionplan · diff · transcript
Sync streaks with last-write-wins, or a CRDT counter?
- planplan · diff · transcript
Weekly summary email: approve the plan?
- mergeplan · diff · transcript
Push reminders: merge #88 into main?
Have a poke around
Three builds at once, one short inbox
Invented data, real mechanics. Agents are building and none of them needs you — except for one architecture call. Open it in the inbox, make the call, and watch a builder pick it up.
Scroll a little further to load the demo
The board
Every piece of work, at the exact stage it's in. Positions aren't decoration — a card's column IS its state.
The inbox
One queue of everything blocked on you, each ask carrying the plan, the diff and the transcript behind it.
Activity
Who's working right now, on what, and what each run cost — live, with every finished run still on the record.
Engineer from anywhere
Your side projects, from your pocket
Answer the architecture call while the kettle boils. The same factory, the same inbox — sized for a phone, so a decision takes a minute wherever you are.
From a beach chair
Scroll a little further to load the demo
The lifecycle
A process, not another chat window
A chat forgets everything the moment you close the tab, which is why it can't be trusted with anything longer than an afternoon. ampm keeps a real database instead: every task is a row with its plan attached, its history intact, and its place in the tree of work it came from. Retries don't overwrite anything — they land beside the last attempt, because you should be able to ask what happened last time.
Plan before you build, test and review before you merge — that discipline wasn't invented here. It's how the best San Francisco engineering orgs actually ship, distilled from fifteen years as an engineer inside them and codified into machinery. You get the institutional knowledge it takes a career to absorb, working for you from day one — and holding the bar even on the nights you wouldn't have bothered.
- Planningyour call
- Implementing
- Testing
- Reviewing code
- Mergingyour call
Questions
Can I run several Claude Code or Codex tasks at the same time?
Yes. Each task runs in its own sandbox and worktree, so as many run at once as your subscription and workers allow — and none of them needs a tab open.
Do I keep control of the architecture?
Yes. When a planner hits a real trade-off it stops and asks, with its recommendation and reasoning. You make the technical calls; agents do the implementation.
Which coding agents does ampm work with?
Claude Code and Codex. ampm runs them on the subscription you already pay for — no API keys and no second bill for tokens — and when one account hits its usage limit, work moves to a backup account without losing anything.
Does ampm merge code on its own?
Nothing merges unless you want it to. Plans, questions and merges wait in your inbox with the plan, diff, pull request and full transcript attached, and you decide which decisions to keep and which to hand off.
Where does the code run, and is my repo safe?
Every task runs in an isolated sandbox on its own worktree, on workers you choose — your laptop, a server you own, or a cloud machine — and can only push to a branch. The output is always a pull request into your repo.
How is this different from chatting with a coding agent?
A chat forgets everything when you close the tab and needs you watching it. ampm keeps every task in a database with its plan and history, runs a real process — plan, implement, test, review, merge — and only asks you when a decision is actually yours.
Can one workspace cover several repos?
Yes. Add the repos agents should change, and add more as read-only context — a design system, an API they call, product docs — so their work matches what's around it. One change can span several repos as one piece of work.
How do I get access?
Join the wait list with Google or GitHub. We're letting people in a few at a time, in the order they joined, and we'll email you when it's your turn.
Early access
Build all of them
We're letting people in a few at a time. Join the wait list with Google or GitHub and we'll email you when it's your turn. Run it on our cloud or on infrastructure you control — your repos, your subscriptions, workers on machines you choose.