Memory
What StoriBot remembers between sessions. There are two memories, and every session reads both before you say a word. Every Stori gets the first; the second grows only when StoriBot decides it should.
Each Stori’s memory
Section titled “Each Stori’s memory”Every Stori has its own memory file, and the builder writes to it as it works, not at the end: what the Stori is and where its data came from, how its pages are built and what they depend on, what you asked for and corrected in your own words, what it decided and why, and what is still open.
Every session starts cold. This file is how the next one picks the work up without re-deriving a decision or contradicting one. When it and the company’s memory disagree, the Stori’s wins.
The company’s memory
Section titled “The company’s memory”One file in the Library: what StoriBot should know about your organisation before any piece of work. What the company does and who it serves, how you refer to things, what is true across every Stori. It is not written on every job. StoriBot adds to it only when it decides something holds for the whole company, and asks you before it moves anything there. You can read and change it in the app under Library, then Memory.
Your part
Section titled “Your part”Durability is StoriBot’s job, and your part is small.
- Say what the work is for and what it should produce, before how it should look. That is what StoriBot asks first, and it is what the memory is built around.
- Correct it in words. “The revenue figure is wrong.” “Always show the quarter before.” What you say gets written down and shapes every later run. What you change by hand on the page does not.