Adopt deep-embed North Star + four-bucket IA with autonomy ladder
Make the deep-embed governance bet actionable for every seam. Open the IA overhaul contract or review the governance state snapshot.
context
The site behaves like a content library: helpful but unable to prove value or protect tone. We're formalizing the operating model so guidance stays respectful, changes are traceable, and outcomes have receipts. Guardrail reference: Narrator tone guide (gr-101).
decision
Adopt the deep-embed model: one North Star line visible on Home and each section landing; organize into navigate, operate, learn, mitigate buckets; enforce the autonomy ladder (nudge-first) and required page opener pattern (one-sentence why, two actions, exit metric).
- Tradeoffs:
- Less copy freedom above the fold in exchange for clarity and measurability.
- Added CI nudges now to avoid costly retrofits later.
- One set of signals capped at eight, so some metrics won't make the cut.
scope
In: all public Band-A pages and annex labs under navigate/operate/learn/mitigate; monthly seam-scoped tags vyyyy.mm-<seam>.
Out: doc engine swaps, paid tooling, staffing changes.
acceptance
- Every covered page renders the opener pattern above first section.
- Frontmatter required:
bucket, north_star_id, guardrail_id, owner, band, date, cta_primary_label, cta_secondary_label, leading_metric, lagging_metric; adddecision_linkwhen scope/metrics change. - Traceability complete:
pr <-> decision/exception <-> guardrail <-> page <-> signal <-> receipt(no orphans). - Signals registry seeded (<= 8 active, each with owner and 30-day SLA).
- Tone allow/ban lists applied; CTA labels pass lint.
- CI budget cold start <= 5 minutes; blocks only on red-lines and broken traceability.
- First seam landing pages shipped with tiny receipts panel.
- All ids and tags are lowercase.
receipts (expected)
- Leading:
m-nav-openup vs baseline;m-lab-pass>= 90% on pilot annex lab. - Lagging:
m-time-to-answerimproves >= 10% in first 30 days;m-decision-hit-rate>= 60% by second tag. - Attach receipts to
v2025.11-governance, then monthly.
freeze and exit
Freeze triggers: annex lab pass < 80%; time-to-answer worsens >= 20% vs baseline; exceptions older than 14 days.
Exit criteria: owner identified; fix applied; proof metric back within 10% of baseline for 7 consecutive days; freeze lifted via noted exception close.