Node Agent: Memory
Last updated: June 10, 2026
Agent Memory, Organization & Project Memory
Give the Scenario Agent persistent context — so it understands your team, your projects, and your preferences from the first message of every new conversation.
Every Agent conversation starts fresh. To carry knowledge across conversations, Scenario gives you three memory layers, each with a different scope and a different author:
Layer | Who writes it | What it's for | Applies to |
Organization Memory | Org admins | Team-wide standards, brand rules, compliance reminders | Everyone in your organization |
Project Memory | Project or Org admins | Project-specific conventions, style guides, references | Everyone on that project |
Personal Memory | The Agent (automatically) | Things the Agent learned about you: preferences, recurring corrections, recurring asks | Just you |
The Agent reads all three at the start of every conversation and uses them together to shape its responses.
How they work together
When you start a new conversation in a project, the Agent assembles its working context in this order (broadest to most specific):
1. Organization Memory: your company's house rules.
2. Project Memory: what's true about this specific project.
3. Your personal instructions: preferences you've written that follow you across every project you're in.
4. Your Personal Memory: what the Agent has remembered about you across projects.
5. Your Personal Memory in this project: what the Agent has remembered about you inside this project specifically.
More specific layers override broader ones. If your Organization says "always render images at 1024×1024" but your Project says "this project ships 1216×832 hero shots" the project rule wins inside that project.
If you've personally told the Agent "I prefer portrait-oriented compositions" in your settings, that overrides both, but only for your conversations.
Organization Memory
Set by: Organization admins.
Applies to: Everyone in the organization, in every project.
Best for:
- Brand guidelines: tone, palette, prohibited subjects.
- Compliance reminders: data-handling rules, NSFW policy, export limits.
- Team conventions: naming, tagging, file-organization patterns.
Set it from: Organization Settings → Assistant.

Example:
> Our brand voice is concise, professional, and never uses exclamation points.
> All hero images must be 16:9 unless a project says otherwise.
> Never include real people's faces without explicit approval from the project lead.
Keep it short and stable. Anything that changes per project belongs in Project Memory instead.
Project Memory
Set by: Project admins (or organization admins from any project).
Applies to: Everyone collaborating on that specific project.
Best for:
- Project-specific style: art direction, color palette, mood references.
- Working conventions: how to name renders, where reference images live, which models the team has standardized on.
- Background context: a paragraph about what the project is, so the Agent doesn't have to be told twice.
Set it from: Project Settings → Assistant.

Example:
> This project is a stylized fantasy mobile RPG. All character renders use the "Aurora" LoRA at scale 0.8 and a 3:4 aspect ratio. Background environments default to dusk lighting, with cool blue-violet hues.
When a team member opens a new conversation in this project, the Agent already knows the visual brief — no re-briefing needed.
Personal Memory
Set by: The Agent itself, automatically, or manually from User Settings.
Applies to: Only you. Other team members have their own.
Best for: The Agent capturing things it's learned about how you work — without you having to repeat yourself.
The Agent writes to its own memory when something useful would help future conversations. Typical examples:
- A preference you've corrected the Agent on more than once.
- A workflow shortcut you use repeatedly.
- A non-obvious choice you've validated (e.g. "yes, the 3:2 aspect ratio is what we want for this kind of asset").
There are two scopes of Personal Memory:
- Across all projects — preferences that follow you everywhere ("I prefer detailed prompts over short ones").
- Inside this project — context only useful in this project ("the character 'Lyra' has silver hair and amber eyes").
Both update silently as you work. You don't have to do anything.
Reviewing or editing what the Agent remembered
You're always in control. From Profile -> User Settings → Personal Memory you can:
- See what's been saved (cross-project memory and per-project memory side by side).
- Edit any entry directly.
- Delete entries that are outdated or wrong.
If you tell the Agent in conversation — "please remember that…" — it will save the instruction to memory. If you say "forget that…" it will remove it.

When to use which
A quick decision guide:
- Should every Organization User know this? → Organization Memory.
- Should every collaborator on this project know this? → Project Memory.
- Is this just a personal preference, a recurring correction, or a shortcut? → Let the Agent save it to Personal Memory.
If you find yourself typing the same correction into chat session after session, that's a sign it belongs in Personal Memory (or, if it applies to everyone, in Project or Organization Memory). The Agent will usually offer to save it for you.
Limits
Each layer has a size budget:
Layer | Size Limit |
Organization Memory | 50 KB |
Project Memory | 50 KB |
Your personal instructions | 20 KB |
Personal Memory (across projects) | 30 KB |
Personal Memory (per project) | 30 KB |
If a layer fills up, the Agent will prune the least relevant entries when it next writes — or you can edit it manually any time.
Best practices
- Be specific. "Use 2-space indentation in JSON outputs" is better than "format outputs cleanly."
- Keep it concise. Shorter memory layers produce better adherence than long ones.
- Don't duplicate. If something belongs in Organization Memory, don't repeat it in every Project Memory — the Agent reads all layers together.
- Audit periodically. Open Settings → Assistant every few weeks and trim what's no longer true. Stale instructions are the most common reason the Agent gives "off-brand" answers.
FAQ
Does Personal Memory share my data with other team members?
No. Personal Memory is per-user. Each person on your team has their own Personal Memory; nobody else can see yours.
Can I turn off Personal Memory?
Yes, from Profile → User Settings → Personal Memory. The Agent will still read Organization and Project Memory, but won't save anything new about you.
What happens to a project's memory if the project is deleted?
All project-scoped memory (Project Memory and per-project Memory for every member) is deleted with the project.
Can the Agent remember something I told it in a chat?
Yes. Just ask it to: "remember that I prefer…" It will save it to Personal Memory. To make it apply to your whole team, ask an admin to add it to Organization or Project Memory instead.
Related
- Organization Settings → Members & Roles — who can edit Organization Memory.
- Project Settings → Members — who can edit Project Memory.
- Profile → User Settings → Personal Memory — review, edit, or delete Personal Memory.