| Main purpose | A decision workspace built for one job: judging whether a startup idea deserves your next weeks. | A general assistant strong in language, reasoning, analysis, coding and long-form or document-based work. |
|---|
| How an idea is evaluated | Clarifying questions first, then the same ten commercial signals applied to every idea. | Open conversation, often with careful reasoning. The framework it uses is whatever your prompt establishes. |
|---|
| Consistency across multiple ideas | Every idea runs through the identical framework, so scores are comparable side by side. | Possible with a saved prompt or Project instructions, if you apply them identically to every idea. |
|---|
| Decision output | One score and one outcome: Build, Explore or Skip, with a stated next step. | A written analysis in the shape you ask for, or an Artifact you can keep and edit. |
|---|
| Assumptions and weak points | Assumptions, risks and a confidence level are named per signal, and the weakest signal is flagged. | Good at drawing these out in discussion. Doing it uniformly on each idea remains your process. |
|---|
| Validation guidance | A validation checklist and a single recommended next action ship with every score. | Can propose tests and next steps when asked for them. |
|---|
| Saved idea portfolio | Ideas are saved as a portfolio you can rank and revisit as they change. | Projects can hold chats, knowledge and instructions. Comparable scores across ideas are not the aim. |
|---|
| Strategy and execution outputs | Marketing, sales, finance and tech outputs plus a full execution plan. | Can produce detailed strategy documents from material and instructions you supply. |
|---|
| Build-ready handoff | A build-ready prompt written for whatever builder you already use. | Can write specs and code, and Artifacts make drafts easy to iterate on. |
|---|
| Best fit | Founders and builders who need a repeatable method for choosing between ideas. | Anyone doing careful reading, reasoning and writing over long inputs. |
|---|