Maturity

Generated 2026-08-07T15:25:56.540757Z · Shadow mode · Policy v1.0

Gate Matrix

Each dimension is an independent gate. The aggregate is the minimum across all six — a factory is only as mature as its weakest gate.

Scope Coverage L3 · sustained

Share of accepted requirements with linked validation evidence.

Value
1.0
Freshness
Fresh
Updated
2026-08-02 00:00 UTC
Source
Requirement-to-test traceability matrices
Thresholds
L0: no matrices · L1: <50% · L2: ≥50% <80% · L3: ≥80% for 14 days
Status
satisfied

Lights-Out L0 · manual

Share of eligible L0/L1 workflows completed without intervention.

Value
—
Freshness
Fresh
Updated
2026-08-07 23:59 UTC
Source
compute_intervention_rate.py output
Thresholds
L0: not measured · L1: <0.50 · L2: ≥0.50 <0.80 · L3: ≥0.80 for 14 days
Status
satisfied

Accepted Work L0 · manual

Share of submitted work units accepted via factory-review PASS.

Value
—
Freshness
Fresh
Updated
2026-08-07 23:59 UTC
Source
compute_daily_scorecard.py + review-verdicts.jsonl
Thresholds
L0: no cache · L1: <0.50 · L2: ≥0.50 <0.80 · L3: ≥0.80 for 14 days
Status
satisfied

Recovery L3 · sustained

Median minutes from failure detection to restored state.

Value
0.0 min
Freshness
Fresh
Updated
2026-08-07 07:59 UTC
Source
compute_recovery_time.py output
Thresholds
L0: no data · L1: >60 min · L2: ≤60 min · L3: ≤30 min for 14 days
Status
satisfied

Provenance L0 · manual

Share of assets with complete factory provenance (recipe, source SHA, license).

Value
—
Freshness
Unknown
Updated
—
Source
check_asset_provenance.py output
Thresholds
L0: no audit · L1: <0.50 · L2: ≥0.50 <0.80 · L3: ≥0.80 for 14 days
Status
BLOCKING

Sustained Evidence L2 · gated

Consecutive sessions with make verify exit 0 and no retroactive defect.

Value
14
Freshness
Fresh
Updated
2026-08-07 08:06 UTC
Source
Git history / verify ledger
Thresholds
L0: 0 clean runs · L1: 1-6 · L2: 7-29 · L3: ≥30
Status
satisfied

Methodology

The factory maturity level is the minimum across six independently-measured dimensions. A factory is only as mature as its weakest measured gate. A weighted average would let a strong dimension mask a failing one — so the aggregate is always the floor, never the average.

Each dimension is scored on a 0–3 scale:

L0 — manual
Dimension is unmeasured, ad-hoc, or has no automation.
L1 — instrumented
Dimension is measured but not gated; data exists but does not block.
L2 — gated
Dimension is measured and enforced as a gate; failure downgrades the aggregate.
L3 — sustained
Dimension is gated and stable at L2+ for the sustained window (14 days).

If any dimension is stale (input older than its max-age) or unknown (no source data), the aggregate is NOT COMPUTABLE regardless of other dimensions' levels. The engine never publishes a level when a dimension is stale or unknown — it reports NOT COMPUTABLE and the dimension's freshness status.

The maturity engine runs in shadow mode for 30 days after the initial deployment. During shadow, the engine computes the level and logs it but does not publish a public maturity badge. This page will display NOT COMPUTABLE until the shadow window completes and the engine is promoted to publication.