ADR-0005 · Unified PARA + Evidence-Guided Taxonomy

We needed a shared vocabulary for the different scales of work (tasks, projects, ventures) and their lifecycle — from spark to production to ongoing maintenance. Two frameworks apply: Tiago Forte’s PARA (folder structure with strict Project = time-bounded + deadline, Area = ongoing) and Itamar Gilad’s idea validation (assess evidence before committing to build). We unified them into a single taxonomy where Ideas enter a Venture Log, get validated proportionally to their scale, become PARA Projects with deadlines, and graduate to PARA Areas when they reach production.

Considered Options

• PARA only, no validation layer: Simple, but doesn’t address “is this worth my time?” — leads to starting projects on gut feeling alone.
• Gilad’s full GIST framework: Designed for product teams with roadmaps. Overkill for a one-person estate. We adopted only the idea validation component.
• Separate systems: Keep PARA for folders and a standalone project tracker for validation. Creates two disconnected views of the same work.

Consequences

• Every use of the word “Project” now means something precise: time-bounded, has a goal, has a deadline. Ongoing things are Areas.
• The Venture Log becomes the lifecycle tracker; PARA stays the folder structure. They connect at clear handoff points (Validated → PARA Project folder; In Production → PARA Area folder).
• Hermes must grill new ideas before scaffolding them. This adds friction by design — friction that prevents wasted effort.
• Tasks (hours of work) stay lightweight: no Venture Log entry, just the mental “is it worth it?” check.