In plain English
This page is part of the technical reference. It keeps the expert detail but starts with a plain-language summary for first-time readers.
- Why this matters: AI risk can come from the whole arrangement, not one obvious model.
- What to look for: data, memory, routes, adapters, tools, evaluators, updates, and rollback paths.
- Technical version below: the expert terminology remains available and is linked through the glossary.
Text Layout and UAI Memory Repair
The v1.13.0 repair fixes visual text collisions where diagram cards could render labels and descriptions as one pushed-together phrase. The repair adds spacing at the HTML source level and at the CSS layout level so the visible layout and the accessibility tree both preserve word boundaries.
What changed
The affected pattern was most visible in small schematic cards such as the plain-language Safe parts, unsafe wholeEach part passes review, but the combined system fails. Open glossary definition diagram. Labels like “Part A passes” and “model check” are now separated as distinct block-level text lines rather than adjacent inline fragments.
Memory strategy
The .uai memory strategy now keeps hot memory concise, stores durable report bodies under /docs, and uses .uai/long-term-memory.uai as a pointer-only ledger. Each source-report pointer names the raw report path, summary path when available, source authority, review status, proof-of-use trail, and checksum.
Report coverage
The repair verifies that all preserved Markdown and PDF source reports under /docs/source-reports/ are represented in .uai/long-term-memory.uai, .uai/report-pointer-index.uai, and per-report .uai/memory/source-reports/ records.
Boundary
This page does not claim UAIX certification, live readiness, or external synchronization. It records local package structure and local validation only.