Product scenarios explain why someone adopts StashBase. They are intentionally high-level and durable: they describe desired outcomes, not screen sequences or test fixtures. Observable shipping flows live in User Journeys.
An engineer starts with requirements, meeting notes, design history, and an existing codebase spread across ordinary local files. StashBase helps an Agent retrieve the relevant constraints, compare options, record accepted decisions in a user-owned document, and check the delivery against the original brief.
The outcome is continuity from requirements through implementation without moving the project into a proprietary workspace.
Related journeys: J02, J05, J06, and J07.
A student or researcher combines an assignment brief, papers, slides, recordings, experiments, and notes. StashBase prepares difficult formats, retrieves evidence across the collection, and keeps accepted conclusions in ordinary Markdown alongside the source material.
The outcome is a traceable project whose sources, decisions, progress, and final report remain connected.
Related journeys: J03, J04, J05, and J07.
A user already has years of notes, PDFs, images, recordings, and earlier Agent output. StashBase indexes selected folders in place, keeps generated representations invisible, and lets built-in or external Agents retrieve the visible sources and write useful new files back.
The outcome is a local archive that becomes more useful over time while the user keeps its existing folder structure and ownership.
Related journeys: J02, J04, J05, and J08.
Project folders often include JSON configuration, fixtures, exports, or partially generated data beside prose. StashBase treats JSON as an ordinary raw source: users can inspect, search, explicitly edit, and save it without requiring validity or silently normalizing its text.
The outcome is one source-of-truth workflow for prose and structured project data.
Related journeys: J03 and J05.
Across these scenarios, work can begin with a question or task for an Agent against an explicit scope. The user may open a source document at any point to inspect evidence or continue the same work beside it, or may complete the task entirely in Chat. Documents are optional working surfaces rather than required entry points. When an outcome needs to persist beyond the conversation, it is explicitly written into an ordinary source file.
The canonical end-to-end product behavior is J10: Turn a local project into durable Agent-assisted work. For Chat-first work, J11: Turn a conversation into a project creates the durable scope before entering that loop. The scenarios above vary the source material and desired result; they do not create separate storage, scope, permission, or Agent models.