What We Tested, What Failed, and What Changed
An evidence-backed account of the Orabuena website tests that failed, what changed because of them, and the workflows that remain unverified.
-
-
Product Decisions and the Tradeoffs Behind Them
The product and website decisions already made while building Orabuena, the tradeoffs accepted, and the evidence or authority still required.
-
Research Lessons, When Supported by Actual Conversations
A public standard for separating verified research observations, interpretations, patterns, decisions, and unresolved questions while building Orabuena.
-
Decide What Needs Your Attention Today
A consequence-based daily attention practice for protecting commitments, unblocking decisions, and advancing the right work across ventures.
-
Keep Client Context Separate Across Ventures
A practical method for keeping client data, promises, permissions, and source evidence attached to the correct venture and relationship.
-
Review Several Businesses Without Mixing Their Priorities
A two-pass review for preserving each venture’s context while coordinating a solopreneur’s shared time, decisions, and capacity.
-
Verify the Work, Not Just the Assistant’s Answer
A practical verification method for matching an AI assistant’s completion claims to persisted, rendered, delivered, and preservation evidence.
-
Give an AI Assistant Clear Approval Boundaries
Define exactly what an AI assistant may prepare, execute, or must return for approval—plus scope, stop conditions, and evidence.
-
What to Delegate to AI—and What to Keep
A practical boundary for delegating preparation and bounded execution to AI while keeping consequential decisions under human authority.
-
What “Done” Should Mean Before You Delegate
Define the outcome, completion state, authority boundary, evidence, and exceptions before delegating consequential work.