My local AI agent remembered things I never said. A reader's security review found it.

Automated excerpt

A 7B model can't reliably tell "the user said X" from "the user asked me to read something that says X". Only the user's own words reach the memory model. self_statements() keeps the clauses where the user talks about themselves. "The user's name is Ada" is dropped when the user only said "My name is Roydon", whatever the model extracted.

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.