What is CuratorNote? A private note workspace for AI-assisted work

What is CuratorNote?

CuratorNote starts from a simple observation: most useful work eventually returns to a record. A decision needs its reasoning, a customer request needs context, a research conclusion needs sources, and an AI-generated draft needs the instructions and evidence that shaped it. A note can hold those relationships without forcing every kind of work into a specialized database. CuratorNote therefore treats notes as durable working records, then connects organization, workflow, and protected storage around them.

The product is not described as an autonomous system that silently runs a company. Its purpose is to give an operator a legible place to capture context and decide what happens next. AI-assisted work can be attached to a record and moved through explicit states, but accountability stays with the person reviewing the result. This distinction matters because a useful assistant can still misunderstand instructions, omit evidence, or produce confident errors. The workspace is meant to make review easier to perform, not easier to forget.

That position also shapes the public vocabulary. A shipped feature is something represented in the current code and available through the active product. An alpha feature is usable in a deliberately limited release. A design principle describes how choices are evaluated, while a roadmap item describes an intended direction that can still change. Keeping those categories visible prevents a reader from treating an architectural ambition as a present guarantee. It also gives early users a more useful basis for feedback: they can evaluate what works now, question the principles, and help determine whether roadmap ideas solve real operating problems.

Who it is for

The initial audience is a solo operator or small team with more context than their current tools can comfortably hold. They may be researching a product, managing a compact customer pipeline, publishing material, learning a technical domain, or coordinating repeatable administrative work. Their challenge is not merely writing text. It is keeping the source record connected to tasks, decisions, revisions, and the assistants that help transform it. CuratorNote is designed around that compact operating environment rather than a complex enterprise hierarchy.

A private workspace is especially relevant when the notes contain product plans, customer history, drafts, or personal research. Privacy does not mean pretending that infrastructure disappears. It means describing what is protected, what remains operationally visible, and which choices belong to the user. The current architecture places protected content encryption in the browser and uses a local working store, while authenticated APIs still handle synchronization and server-visible metadata. The detailed boundaries are covered in the dedicated security-model guide and the reviewed privacy policy.

CuratorNote may be a poor fit for a team that needs mature enterprise administration, a finalized compliance program, or fully autonomous execution today. It is also not intended to replace specialist systems when a regulated or high-impact decision requires dedicated controls. The alpha is a focused environment for people willing to test a note-centered model and inspect its boundaries. That narrower promise is more helpful than presenting an early product as a universal operating system.

How it works

A user first authenticates with email OTP or Google OAuth where configured. After authentication, protected workspaces pass through a master-password experience that creates or unlocks encryption material in the browser. The app maintains local-first data in IndexedDB and exchanges authenticated records with the service. Content designed for protection is encrypted before it leaves the browser, while identifiers, timestamps, request information, and other operational metadata may remain visible to the server. That division is more precise than using one broad privacy slogan.

Within the workspace, notes provide stable context. Categories, tags, media, and related views help organize that context, while the product direction adds work surfaces that reuse the same records. A Kanban workflow is a natural first extension because its states are understandable to both people and software: work can wait in a backlog, become active, move into review, and finish with a recorded outcome. The important design choice is that the board points back to evidence and notes instead of becoming an isolated list of opaque tasks.

What is shipped today

The current application includes authenticated note management, categories, tags, media-related routes, settings, notifications, and an agents area within the active product shell. Browser authentication uses an HttpOnly session cookie, and direct API access can use bearer credentials. Protected routes incorporate the encryption-key and master-password gates. The local application cache uses IndexedDB, and the Cloudflare Worker provides authenticated endpoints backed by the infrastructure bindings required by the implementation. These are code-verifiable foundations rather than promises about future outcomes.

Availability should still be understood in the context of a private alpha. The marketing site does not claim that every proposed workflow is broadly released or complete. Pricing is invite-based or listed as TBD, onboarding is founder-led, and some routes represent an evolving product surface. Comments are currently unavailable, for example. This kind of status labeling is intentional: people evaluating a young product need to distinguish a working foundation from an experiment, a design principle, or a roadmap item.

Roadmap direction

The roadmap extends a note-centered system into visible collaboration between people and AI agents. The likely pattern is not a hidden automation pipeline. It is a sequence in which an operator selects context, defines a bounded task, inspects output, records corrections, and decides whether work is done. Kanban can make those transitions legible. Other work apps, such as lightweight learning or customer-memory views, can then reuse the same note records rather than creating separate silos with inconsistent copies.

Roadmap language is a statement of direction, not a delivery commitment. Details can change after testing reveals that a workflow is confusing, unsafe, or insufficiently useful. Agent access needs explicit scope, logs need to support human understanding, and failure states need a place in the interface. The NIST AI Risk Management Framework offers a useful public vocabulary for thinking about governance, measurement, and management, but adopting that vocabulary does not itself certify a product or remove the need for product-specific judgment.

Privacy and ownership

CuratorNote's security approach begins with browser encryption for protected content and a local-first copy for active work. This design reduces the amount of readable content that ordinary server storage needs to handle, but it is not a claim that every aspect of a request is concealed. Authentication identifiers, server-visible metadata, network information, object sizes, and timing can remain available to infrastructure. A compromised browser, malicious extension, exposed password, or intentionally exported file can also bypass the intended protection boundary.

Ownership includes more than encryption. People need understandable exports, clear deletion behavior, predictable access boundaries, and honest status information. The reviewed privacy policy supplies the approved operational and legal details from the canonical release record. A trustworthy product should keep those disclosures aligned with implementation changes. Readers who need the mechanics can continue to How CuratorNote works, the browser-encryption security model, or the broader security overview before deciding whether the alpha fits their needs.