# L1 — Rigidity candidate sensor table (first fill)

**Status:** first candidates filled from VoteView / Congress / E2 evidence · research-only · **no Rigidity Index** · no scores  
**CSV:** `data/lanes/L1/rigidity_candidate_sensors_v0.csv`  
**Research job:** `lane-L1-system-rigidity` · Hub `b73c292a` · ACK `b8c1c81e`  
**MODEL_CHANGE=NO**

## Rules (unchanged)
- Do **not** build a Rigidity Index.
- Disposition ∈ {KEEP_CANDIDATE, MODIFY, INSUFFICIENT}.
- Trend direction is **irrelevant** to source selection (`trend_irrelevant_to_selection=YES` on all rows).
- Claude C1 discriminant / Codex corrections: numerical cutoffs remain **PROPOSED**, not acceptance gates.
- Overlap with S1/E2 must be explicit before KEEP.

## Filled candidates

| ID | Dim | Construct | Source | Disposition |
|----|-----|-----------|--------|-------------|
| L1R-A1 | A within-bloc cohesion | Within-party NOMINATE SD (chamber×party×Congress) | VoteView / E2-N1 | **KEEP_CANDIDATE** |
| L1R-A2 | A within-bloc cohesion | Brookings party-unity vote % | Brookings Vital Stats / E2-N2 family | **MODIFY** (coding/chamber predeclare; partial dependence w/ VoteView) |
| L1R-B1 | B between-bloc separation | Between-party median center distance | VoteView / E2-N3 | **KEEP_CANDIDATE** |
| L1R-B2 | B between-bloc separation | Between-party mean center distance | VoteView robustness twin | **MODIFY** (not independent of B1) |
| L1R-C1 | C boundary permeability | CBS cross-party **offer** rate | CEL CBS / E2-N4 | **KEEP_CANDIDATE** |
| L1R-C2 | C boundary permeability | CBS cross-party **attract** rate | CEL CBS / E2-N4 | **MODIFY** (pair with C1; forbid composite) |

## Coverage notes (metadata)
- VoteView dispersion/distance: congress 25–118/119; continuous; House primary for B1.
- CBS offer/attract: congress 98–118 only (shorter; binds C-dimension overlap with A/B to 21 congresses).
- No S1 mass series used as elite rigidity sensors (overlap would be cross-layer, not within-elite).
- **INSUFFICIENT** not used this fill — all six rows have landed series; MODIFY flags protocol gaps.

## Explicit non-claims
No Rigidity Index. No E2/L1 score changes. No composites. No regressions. C1 operational discriminant still pending Astra/Claude freeze.