The new 24-slide guide and
PDF teach Tara's five duration buckets,
ten context tags and eight lineage records, followed by the implemented lookup,
quiet-hours, persona and validation paths. The
authoring record preserves the
source review, sequence, design decisions and unresolved application boundary.
Implementation and presentation/audio base:
096fb902ddce8134ec356a93574d581e0ae9b29e;
primary documentation pin: 1ceae52c7896a59ac74a4429d410f3218f0b91c6.
The library now has 74 guides, 878 unique slides, 901 guide-slide instances and 427 cited paths, all with complete narration. The 24 new tracks add 1,484.102 seconds (24.7 minutes); all 854 baseline scripts, manifest entries and MP3 bytes remain exact. Three reviewed assignments advance the Tara handbook source to 13 reviewed, 2 structural and 15 unassigned units. Global coverage is 91 reviewed, 10,018 structural and 103,235 unassigned units, so the full V1 program remains active.
Environment records the browser, runtime, speech package and resource-admission details used for this delivery.
Evidence#
- Fourteen source cases execute the complete pinned feature modules and their dependencies: exact catalog records, every duration and local-hour boundary, the 12-minute availability edge, catalog-order signal results, quiet-hours filtering, lineage/persona branches and validator failure paths. A bounded direct-symbol search found no application caller for the three helpers. Nine-file continuity binds the owning modules and specs across the pin and integrated repository baseline.
- The final browser report passes 172 checks, retains 96 desktop/mobile and embedded captures, runs 72 embedded axe audits, exercises four enlarged-text cases and rejects four evidence removals. There are no fit findings or unexpected browser errors.
- Focused center review passes all 14 checks across the 74-guide catalog, proves 73 guide HTML files unchanged from the presentation baseline, retains 48 selected-guide workspace captures, exercises both audio boundaries and exports all 24 PDF pages. Exact visible text/order and optical review separately cover the PDF.
- A fresh
@tara/featuresrun passes, and the settled presentation suite passes 147 tests. Scoped narration decode/pause, generated-artifact freshness, teaching coverage, ESLint, Ruff, formatting and local-link checks pass. Speech reuse binds prior entries, scripts and bytes; it does not claim a new normal-speed human listening session.
Corrections and limits#
The first 320 px review found one unbreakable LineageTaxonomyEntry specimen
heading. The readable label “Lineage taxonomy entry” removes the document
overflow while the exact source type remains part of the source review. A second
draft reached the center playback check before narration existed; the final run
includes all audio and passes. The first settled unit suite also identified the
cumulative rollback-slide allowlist and earlier Tara source-count expectation;
both guards now include this reviewed increment, and all 147 cases pass.
The worked examples use declared inputs and executed library behavior. No live member, clock, calendar, sensor, content assembly, recommendation delivery, lineage authority, rights holder, clinician or application launch participated. The direct-symbol search can miss wrappers or alternate algorithms. Human comprehension, cultural/rights acceptance and normal-speed listening remain unperformed.
Run
python3 docs/presentations/presentation-center/verification/v1-tara-context-lineage-2026-09-09/verify-receipt.py
from the repository root after checkout. Later Presentation Center edits
invalidate this frozen receipt; use the delivery revision for historical
verification. The full V1 program remains
active.