中文

Open-source Personal Agent State Layer

Memory isn't enough.

Trusted personal state for every AI.
Keep important facts, decisions, and progress structured, current, and traceable.

Try the demo ▼
Change it yourselfIn-browser demo. No model involved.
codexChat
>
▸▸ Auto-sync on
Problem

Problem & Aha

Your AI remembers. But is it still right?

It remembers everything you said, but not which parts have changed. So you keep correcting it.

Illustrative scenarios of what can go wrong when an AI relies on scattered history. Not every memory product makes these mistakes.

Remembering isn't the same as staying current.

Less scattered memory. More reliable state.

Keep important information organized and updated, so AI can pick up where you left off — across conversations, tools, and agents.

Memory1 / 5

Memory is the input.

What your AI keeps is a few scattered lines

Launch date, payment progress, technical approach… scattered across one conversation after another. Every line is remembered. None is marked as still true.

AI MEMORYScattered
“Launch next Friday”“Still need payments”“Going with Option B”“Design approved”“Launch email not written yet”
? Which still count
State2 / 5

Less scattered memory. More reliable state.

Important information shouldn't be scattered memory

Facts, decisions, plans, progress: each has one clear current state.

SAVESLOT · Product launchv1
Launch dateNext FridayCurrent
PaymentsPendingCurrent
ApproachOption BCurrent
DesignApprovedCurrent
Launch emailTo doCurrent
State3 / 5

Current, not stale

When things change, your AI should know.

Say it once. The old value is marked “superseded”, and what your AI reads next is current.

Old versions are never deleted; view or roll back any time. The above is the target experience, to be validated in the product.

SAVESLOT · Product launchv2

“Launch moved to next Tuesday.”

Launch dateNext FridaySuperseded
Launch dateNext Tuesday★ CURRENT
PaymentsPendingCurrent
ApproachOption BCurrent
DesignApprovedCurrent
Launch emailTo doCurrent
Continuity4 / 5

Continuity

Pick up where you left off.

Start a new session and your AI reads the current state first, then continues from where you stopped. No re-briefing.

SAVESLOT · Product launchv2
Launch dateNext TuesdayCurrent
PaymentsPendingCurrent
DesignApprovedCurrent
Launch emailTo doCurrent
Thursday · new session

Launch is set for next Tuesday and the design is approved. Next: integrate payments, then write the launch email.Read v2

Continuity5 / 5

Trusted state across conversations and AI agents.

Not tied to any one AI

Different AIs read and write the same state of yours instead of each keeping its own. Switch models or tools; your progress stays.

Illustrative flow of the target product experience. What ships at launch is what counts.

YOUR SAVESLOTOne copy · v3
Launch dateNext TuesdayCurrent
PaymentsDone★ CURRENT
DesignApprovedCurrent
Launch emailTo doCurrent
CodexWrote v3Claude CodeReadChatGPTReadClaudeReadCursorRead

Trusted State

What makes state trustworthy?

Trusted doesn't mean guaranteed correct, and SaveSlot doesn't claim to verify the real world for you. It means an agent can tell, within a clear scope, whether a piece of information is current, where it came from, whether it was confirmed, and whether it may use it.

These are design goals, not shipped features. What ships at launch is what counts.

NOW

Current

Tells the current value from past values and unconfirmed updates, and records when each took effect and was last updated.

SRC

Provenance

Records whether you confirmed it, an agent inferred it from a chat, or a third party synced it. An inference never passes as fact.

ONE

Consistent

When edits, rollbacks, or contradictory conclusions collide, explicit rules pick the valid version, or ask you to confirm.

KEY

Controlled

Reading, writing, correcting, and deleting are authorized separately. No AI gets full access by default.

LOG

Traceable

Keeps the change history and audit trail needed to explain why a value is what it is, and to restore an earlier one.

Your data stays yours

Your state doesn't belong to any AI platform. Visible, portable, deletable.

{}

Open-source core

Public code. The self-hosted version is fully featured.

↓

No platform lock-in

State is independent of any AI platform. Uninstalling a plugin never loses a save; export and delete are always free.

◐

Least privilege

Only AIs you authorize can read or write. Each plugin declares what it touches and whether it goes online. Revoke any time.

✓

Public audits (planned)

We plan to publish SECURITY.md and a threat model. There are no third-party audit findings yet.

How we measure it

How we measure it

“Remembers it right” can't rest on a slogan. These are the six metrics we will measure. They are targets to validate, not results.

■ We'll publish the method and reproducible results at launch. Until then, no numbers here.

At launch

Stale-state rate

Freshness

How much of the state an AI reads is no longer the current value.

At launch

Superseded-state misuse rate

Misuse

How often an AI acts on a value that has already been superseded.

At launch

Corrections and repeated explanations

Repeats

How many times you have to correct your AI or explain the same thing again.

At launch

Conflict-detection accuracy

Conflict

When new and old information contradict, does it catch that and pick the right version, or ask you?

At launch

Provenance coverage

Provenance

How much of the state can say where it came from and why it changed.

At launch

Cross-session continuation rate

Continuity

After a new session or a different AI, can work pick up where it was?

Plugins

One core. One plugin per job.

Plugins share one system of versions, permissions, provenance and rollback. Add a second one without learning it again.

01

Project Keeper

At launch

Decisions, tasks and handoffs across sessions and coding agents.

02

Body Log

At launch

Meals, workouts, weight and a weekly recap. No medical diagnosis.

03

Goal / Decision Tracker

Example

Goals, decisions and why they changed. A demo of how to build a new plugin.

--

Build your own plugin

Open

Job search, budget, study, family… build your own with the Plugin Starter. No core changes needed.

Waitlist · step 2

What should your AI always remember?

Optional; skip it if you like. We read every answer and use them to decide what to build first.

Used only to tell you when we launch and to understand what you need. Privacy notes

PREVIEW · after launch, you hand your AI this prompt to connect

Planned pricing

Pick your plan

■ Planned pricing. May change before launch.

■ Export, delete, correct and manage permissions: free on every plan, forever

Click a card or use ← → to pick a plan, then join the waitlist

FAQ

Questions you might have

Didn't find your answer? , I read every one. Or ask on Discord directly.

Get on the list

Make your AI remember it right

Once we launch, you say a change once and your AI starts from the latest state every time. Leave your email and we'll tell you the moment we launch.

Back to the demo
AIDEN · FOUNDER

Me again. You made it to the end, which is more patience than my own AI has. Tell me what you didn't get, where you think I'm bragging, or what made you want to swear. I read every message myself, and I remember them right. To rant in person, find me on Discord.

GET saveslot.ai/state.json200 OK

Same site. This is what an agent reads.