Open-World Narrative · Tooling

V5 authored case corpus gate

A tooling document in the Open-World Narrative library.

0sections1 minread

On this page

This gate is the source-of-truth boundary between V5's launch-case inventory and the human-authored, playable corpus required by V8.

The repository currently has 25 Vice Squad inventory rows plus separate 20-pack and 8-case cold-case skeletons. Those records prove intended counts and runtime identifiers; they do not contain complete case truth, prose, recorded voice-over, playable assets, rights receipts, or validated deduction graphs. The corpus manifest therefore declares acquisition-required and contains no case records. This is intentional: planning metadata must not be promoted into human-gold content.

Run an honest inventory audit:

bash
python3 V5/tools/authored-case-corpus/validate-authored-case-corpus.py

Use the strict release gate before claiming V5-002 or consuming the corpus as V8 regression gold:

bash
python3 V5/tools/authored-case-corpus/validate-authored-case-corpus.py --strict

Strict mode requires exactly five cases for each launch desk and at least five held-out cases. Every case must bind canon, culprit, means/motive/opportunity, fair-play clues, eliminable red herrings, one acyclic solution graph, timing, assets, prose, recorded VO, outcomes, human author identity, source/asset rights receipts, primary locale, tone/style, playable runtime evidence, and contamination controls. Every referenced repository artifact must exist.

Audit mode exits successfully while reporting ready: false; strict mode exits non-zero for the same incomplete manifest. This lets routine repository checks preserve an accurate inventory while keeping the product/release claim fail-closed.