Useful tools, built from
real requests.

16 tools, plus your own workflows

The beta provides 16 browser-based utilities, keyword-based intent search, local favourites, recently used tools, reusable text workflows, request submission and feedback.

Existing tools process their input on the user’s device. A request is a separate workflow: users describe a desired transformation and can follow the request status returned by the server.

A personal workflow combines up to eight supported steps. Users can preview the result, save up to 25 named workflows on their device, edit and run them again, and export or import the recipes as JSON. Execution inputs and outputs are not saved with a recipe.

The library is intended for practical text and data tasks. It does not require a subscription or payment to try the current beta.

A constrained, reviewable loop

1. A user submits a plain-language request, optional input and expected-output examples, and a repeat frequency.
2. When an AI provider is connected, the server drafts a specification using allowed text-processing operations.
3. The draft is validated against supported operations and checked with examples.
4. A developer reviews the draft before approving publication.
5. Only approved tools become reusable entries in the library.

The current loop composes supported operations. It does not run arbitrary generated source code, promise to fulfil every request, or publish drafts without developer approval.

Our team and roadmap

Moadeck was founded in December 2025 in South Korea. Its dedicated development team is building the bootstrapped utility platform.

We are testing which tasks users repeat, which requests recur, and whether reviewed tools are useful enough to revisit. The product collects direct feedback. Optional beta research measures consenting visitors’ feature completion and repeat usage without collecting utility inputs or workflow names. Demo activity is excluded from those research metrics.

Planned work includes better reuse of similar requests, demand-informed prioritisation, broader transformations and more capable AI-assisted development. We will expand the supported scope while retaining example checks and developer review before publication.

Data principles

For the existing browser utilities, input and output are processed locally. Favourites, recent tool identifiers, request identifiers and saved workflow recipes are stored in the browser. A recipe includes its name, steps and any literal replacement strings. Execution inputs and outputs are not persisted; utility input text is not stored as usage analytics.

Submitted tool requests, feedback and optional contact information are sent to the service. Consented requests may be passed to the configured AI provider for draft generation. Users should exclude secrets and sensitive information.

Aggregate usage metrics do not include search phrases or utility input contents. Submitted contact and feedback data have a maximum retention period of 180 days. Contact us to request access or deletion.