Model

The developer's index into where LCLS-MFX content lives, and the architectural gap register this exercise produced. Design-phase.

MFX is a documentation-and-descriptor scaffold: it exists as the descriptor and docs below, not yet as registered events or integration scenarios. This page points to where each piece lives, then records the gaps the exercise found.

Kind	Where	Notes
Beamline descriptor	deployments/lcls-mfx/beamline.yaml	the device walk, with the pcdshub-derived EPICS PV prefixes; source of the generated Source page
Site descriptor	deployments/slac/site.yaml	the SLAC facility surface; MFX is its only beamline, with XFEL practices carried pending
Catalog Family	catalog/catalog.yaml	no Family graduated at the time. MFX reuses catalog Families, including the graduated Transfocator (a CRL focusing optic), and carries loose families (FluxMonitor, Diagnostic, Laser reused). It introduced EmissionSpectrometer, which has since GRADUATED into the catalog (ISS 2nd sighting, SPEC-1)
Catalog Capability / Method	catalog/catalog.yaml	none added; serial crystallography, pump-probe, and emission-spectroscopy Methods are deferred (the catalog tomography Methods do not fit an XFEL)
Catalog Model	catalog/catalog.yaml	none bound; pcdshub names hardware (Dectris, Rayonix, the von Hamos) but no part is procured
Equipment Assets	not yet registered	the Inventory is the planned shape; no scenario registers MFX Assets yet
Trust / governance	not yet instantiated	see Governance, including the pump-probe laser Clearance

The headline: the families fold, the gaps are architectural

I03 graduated a device Family (Goniometer). MFX coined exactly one new family and reused everything else, and that is the finding. Of MFX's full device set, one type had no CORA Family, the von Hamos emission spectrometer; it introduced EmissionSpectrometer, which has since GRADUATED into the catalog once NSLS-II ISS (8-ID) earned the second sighting (SPEC-1). Everything else reuses an existing Family: the offset mirrors fold into Mirror, the solid-Si attenuators into Filter, the JAWS into Slit, the pulse picker into Shutter (PULSE-1), the profile imagers into Scintillator + Camera, the intensity-position monitors into FluxMonitor + Diagnostic, the channel-cut into Monochromator, the lens stacks into the graduated Transfocator catalog Family, the EventSequencer into TimingController, the pump-probe laser into the loose Laser (the 4-ID precedent), and the area detector into Camera. Each fold was reviewed against coining a synonym and rejected.

So the device taxonomy generalizes from storage rings to an XFEL almost untouched. What does not generalize is the acquisition ontology. That is the product of this exercise, recorded next.

Deliberately not here yet (the architectural gap register)

These are the parts of MFX this scaffold leaves out on purpose. Unlike the open questions (facts the LCLS team owns), each of these is a CORA scope decision: a shape the model does not yet have, with the seam it would extend named. None is built speculatively; an XFEL is the trigger that would justify the work.

• Per-shot, pulse-ID-tagged event DAQ (DAQ-1). The load-bearing gap. CORA's acquisition is a single-detector poll-to-Done loop (apps/api/src/cora/operation/acquisitions.py) plus a sub-Hz scalar observation logbook with no pulse-ID key (apps/api/src/cora/run/aggregates/run/entries.py). An XFEL collects a free-running stream of per-shot frames correlated by fiducial at beam rate. The Run-as-provenance-envelope survives and the per-shot data plane lives in psana (CORA references a Dataset, as it does for reconstructions via ComputePort), but representing a DAQ run as an actuation is a new event-stream axis. Its shape is sketched as a forward-looking design note in CORA's design memory (gated, not built).
• Beam-synchronous event-code timing (TIMING-1). The EventSequencer plays a sequence of [beam_code, delta_beam, delta_fiducial, burst_count] lines that gate acquisition at beam rate. CORA's TimingController Family carries the device, but "acquire on event-code N at rate R, burst B" has no typed parameter home; today it would be opaque setpoints.
• Femtosecond pump-probe synchronization (LASER-1). The optical laser and the FEL are two synchronized timing domains (the lxt_ttc SyncAxis holds a ~50 fs deadband; the timetool corrects residual jitter). CORA's PartitionRule is single-domain spatial math; a cross-timing-domain synchronization is a relationship it cannot express. The laser device itself folds (loose Laser, 4-ID precedent); the sync is the gap.
• One switched FEL source feeding co-equal instruments (TOPO-1). One linac and undulator line serve many instruments, beam routed one at a time by the transport mirrors. CORA models each beamline as a root Unit owning its source; a shared, switched source feeding co-equal Units has no home except the Supply("PhotonBeam") seam, and the routing state ("which instrument has beam now") is new.
• Attenuator transmission solver (ATT-1). The solid-Si attenuators solve a foil combination for a requested transmission, energy-dependent (the AttBase solver). CORA's Filter covers the discrete selection; the solve is the deferred Attenuable + SolverReference leg (apps/api/src/cora/operation/_partition_rule_eval.py defers SolverReference evaluation). MFX, with the same focus solver on the Transfocator lens stacks, is the rule-of-three trigger.
• Computed device-state to path-transmission lightpath (LIGHTPATH-1). pcdshub's lightpath walks the z-ordered beam path and computes path-level transmission and the first blocking device from each device's inserted / removed state. CORA already has the static z-ordered walk and the location-not-identity discipline; only the dynamic computed half is deferred (the passive-beam-path tier). lightpath/path.py is a ready precedent.

What is deliberately not here yet (modelling, as at the other exercises)

• New Capabilities / Methods and vendor Models. MFX earns no catalog change; the XFEL recipes are carried pending on the SLAC Practices. No catalog Model is bound.
• The von Hamos as a graduated Family. EmissionSpectrometer GRADUATED into the catalog once NSLS-II ISS (8-ID) earned the second sighting (its Johann + von Hamos XES / HERFD spectrometers); MAX IV Balder (SCANIA-2D) is a third near-sighting. The residual open question is whether each analyzer crystal is a child Asset (SPEC-1).
• Sample delivery and the Subject custody thread. The liquid jet / fixed target is endstation-specific and deferred (SAMPLE-1); no Family is coined.
• Integration scenarios. No test_lcls_mfx_*.py registers MFX Assets. Hard-registering a design-phase, off-roadmap, XFEL beamline would commit speculative structure.

The 2-BM Model page shows the by-kind index a fully-modelled deployment carries.