Skip to content

Deep-analysis research lanes

SCPN Quantum Control contains a substantial collection of analysis and gauge modules. They do not all have the same maturity, evidence, or relationship to differentiable control. The BL-84 registry makes those differences explicit without turning importability into a product or scientific claim.

The committed snapshot covers 74 modules: 29 research lanes, 29 prototypes, and 16 product candidates. “Product candidate” is a software-maturity label, not productisation. Every row explicitly denies product, control, and publication grants from registry membership alone.

Read a row

Each row has four required classification fields:

Field Meaning
module Exact import path under scpn_quantum_control.analysis or scpn_quantum_control.gauge
maturity research, prototype, or product_candidate software maturity
diff_hook No relationship, diagnostic use, an evidence-requiring candidate, a bounded existing composition, or an owner-gated deferred route
claim_status research_only, diagnostic_only, evidence_bounded, or a fail-closed refuse_only interface

Rows may also name a promotion target and a repository evidence pointer. An empty evidence list is intentional for research-only and diagnostic-only code. An evidence_bounded row cannot be constructed without an evidence pointer.

from scpn_quantum_control.analysis import get_research_lane

lane = get_research_lane("scpn_quantum_control.analysis.qfi")
print(lane.maturity.value)          # prototype
print(lane.diff_hook.value)         # candidate_requires_evidence
print(lane.promotion_targets)       # ('BL-50:planned',)
print(lane.registry_grants_control) # False

Unknown modules raise KeyError; the API never synthesises a permissive default.

Promotion routes

The registry records route status rather than flattening every relationship into “promoted”:

Route Current meaning
BL-50 geometric QFI control planned; QFI modules remain candidates requiring separate evidence
BL-54 DLA/topology control complete; selected DLA assets are composed by BL-54's separate bounded package and evidence
BL-72 TCBO observer deferred-owner-gate; the tier-C route remains blocked without owner override
BL-80 adaptive FIM complete; the feedback and custody modules point to their separate committed evidence

These links do not transfer the target's status to unrelated modules. In particular, BL-54 and BL-80 completion does not promote the full analysis catalogue.

Inventory gate

assert_research_lane_inventory() discovers ordinary top-level Python modules in the installed analysis/ and gauge/ packages. It compares that set with the immutable reviewed rows. A new module without a row, or a row whose module was removed, raises RuntimeError and names the exact drift.

from scpn_quantum_control.analysis import assert_research_lane_inventory

report = assert_research_lane_inventory()
assert report.passed
assert not report.missing_modules
assert not report.orphaned_records

Only package __init__.py files and the registry implementation itself are excluded. This keeps the gate focused on research lanes while avoiding a self-registering governance row.

For CI or review, verify the committed deterministic artefacts:

python scripts/run_research_lane_registry.py --check

The command performs inventory validation before comparing bytes. It does not load credentials, contact a provider, submit QPU work, or execute hardware. The committed human-readable matrix is data/research_lane_registry/evidence.md, paired with the machine-readable JSON under the same directory.

Claim boundary

Registry membership is catalogue metadata only. It does not grant productisation, differentiability, control, hardware validity, advantage, criticality, topology certification, consciousness interpretation, clinical interpretation, or a publication claim.

For the six individually reviewed experimental theory hooks, also consult the Theory-Hook Promotion Matrix. For complete symbols, exceptions, and serialization contracts, use the Research-lane Registry API.