| title | Declared outcomes example |
|---|---|
| type | guide |
| status | draft |
| owner | Ada Example |
| created | 2026-09-17 |
| last_updated | 2026-09-17 |
| source_of_truth | false |
| sensitivity | public |
| source_ids |
All people, sources, values and assessments are synthetic. The fixed day makes
checks reproducible; this fixture supplies no evidence of product effectiveness.
The restricted [[00-context/outcomes/o-001|Outcome O-001]] declares a baseline
of 1.000, target 3.000 and observation 2.000 items for approved D-003.
Both historical samples pin [[00-context/claims/c-002|claim C-002]], whose quote
binds exact captured fixture bytes. These checks do not compare numeric meaning,
prove causality, authenticate Ada Example or authorize any execution.
A separate [[06-decisions/d-005-keep-measured-outcomes-separate-from-approvals-and-causal-claims|draft format decision]] records the change. Existing approved reasoning and receipts were not rewritten. Public exports omit restricted declarations and claims. Base64 capture storage is reversible and is not encryption.
From this unreleased source checkout:
uv run whykit outcome verify O-001 --root examples/outcomes --max-sensitivity restricted --today 2026-09-17 --json
uv run whykit knowledge view --root examples/outcomes --max-sensitivity restricted --json
uv run whykit lint examples/outcomes --strict --today 2026-09-17outcome-input.json is the actual input used to create O-001.
A new create preview allocates the next O-ID; it does not modify O-001.
See docs/outcomes.md in the source checkout for preview/apply and measurement limits,
and docs/knowledge-v2.md for private bundle/restore commands.
A bundle preserves raw bytes but does not transfer Git history or establish a
trusted approval baseline. Safe export/restore currently require POSIX support.