Feature

Agents that remember everything.

Every project gets its own database and filesystem. Agents accumulate knowledge across runs, share findings with each other, and get smarter the longer they operate.

What you get

Project database

Every project gets its own PostgreSQL database. Agents store and query real structured data. Not a simple notepad, a full database.

Project filesystem

A persistent filesystem scoped to the project. Agents read and write files that survive across runs: logs, exports, drafts, and generated reports.

Accumulating knowledge

Agents build up knowledge over time. Each run can read what previous runs discovered, making agents smarter the longer they operate.

Shared across agents

All agents in a project share the same database and filesystem. One agent can store findings that another agent uses later, without manual wiring.

Queryable history

Agents can query their own run history and stored data to make better decisions. They remember what they found last week and act on it today.

No external dependencies

Memory is built into the platform. You don't need to set up a separate database, object store, or vector DB. It just works.

How it works

When you create a project, a dedicated PostgreSQL database and filesystem are provisioned automatically.

Agents read and write data using the permissions you grant. No setup required.

Data persists across runs. An agent can store competitive intelligence on Monday and reference it in a Friday report.

All agents in the project share the same memory, enabling coordination and knowledge transfer between specialized agents.

Ready to get started?

Set up your first project in minutes. Connect your tools, configure an agent, and let it run.