diff --git a/.beads/issues.jsonl b/.beads/issues.jsonl
index 02775f0..79f73c0 100644
--- a/.beads/issues.jsonl
+++ b/.beads/issues.jsonl
@@ -1,3 +1,4 @@
+{"id":"homemaker-py-mz5","title":"Python native fitness evaluation (port urb-fitness.pl)","description":"We need a Python implementation of the urb-fitness scoring tool that is faithful to the Perl oracle (urb-fitness.pl / ProgrammeDriven.pm). This is the 'native fitness' component identified in DESIGN.md §6 as gating topology search at scale — the oracle requires a subprocess+file roundtrip per eval which is too slow for large populations.\n\nThe native fitness must reproduce all scoring terms from the Perl source:\n- size, width, proportion (per-space Gaussian scoring)\n- adjacency, access/inaccessible, crinkliness, perpendicular\n- level, staircase volume/count, public access\n- circulation \u0026 outside ratios, min internal area\n\nSource of truth: /home/bruno/src/urb/lib/Urb/Dom/Fitness/ProgrammeDriven.pm and the Storey/Building/Leaf/Base submodules.\n\nValidation target: match oracle scores on the programme-house corpus (35+ .dom files) to within the ~3.7% gap documented in homemaker-py-gpx.","status":"open","priority":1,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-06-15T22:18:06Z","created_by":"Bruno Postle","updated_at":"2026-06-15T22:18:06Z","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-40i","title":"Investigate cf0b8a77e8b2325f ~18% raw_value discrepancy (py lower than oracle)","description":"For prefix cf0b8a77e8b2325f: oracle=1.079112e-03 py=9.133243e-04 ratio=0.8464 (python is ~18% too low). debug_nfails shows py n_fails=5 oracle n_fails=5 (same failures), stair_fits=[1.3145] in python, building_factor=0.1104 (vs oracle's implied ~0.1303). The discrepancy is in raw_value (py=11837 vs oracle implied ~13975) or possibly building_factor. Need to check: (1) per-leaf quality values (crinkliness, area_outside, access) via debug_quality.txt; (2) whether the stair corners differ (cf/rl: py=[2,3] perl=[2,3] — SAME, so corners ok); (3) any quality term not yet ported or computed differently. Run debug_quality.py and compare per-leaf contributions.","status":"closed","priority":1,"issue_type":"bug","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-13T18:08:22Z","created_by":"Bruno Postle","updated_at":"2026-06-13T19:54:04Z","started_at":"2026-06-13T18:12:23Z","closed_at":"2026-06-13T19:54:04Z","close_reason":"Investigation complete: traced 18% discrepancy (cf0b8a77) through entrance corner logic and weighted path length bugs, both now fixed in w1e.","dependencies":[{"issue_id":"homemaker-py-40i","depends_on_id":"homemaker-py-hgg","type":"blocks","created_at":"2026-06-13T19:08:30Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":1,"comment_count":0}
 {"id":"homemaker-py-w1e","title":"Port Perl entrance-corner logic into Python stair-fit (ca/cb parity)","description":"Perl's check_stair_fit (Leaf.pm:104-142) adds entrance edge corners to corners_in_use before computing stair_fit. Python's process_storey does not. For ca9e80c5c1502f10 and cb93a2d2de7f5d37 the oracle stair leaf 'lr' has corners [1,2,3] (from perl_bf.pl) but debug_corners.py Perl-compatible trace gives [2,3] — the extra corner 1 comes from Entrances(graph)-\u003eBoundary_Id logic. Need to: (1) port dom.Entrances() — returns {leaf_id: boundary_id} for the best-entrance leaf at ground level (Entrances() returns {} if level\u003e0); (2) port leaf.Boundary_Id(side) — returns the node sharing that edge; (3) in fitness.process_storey, after stack_corners_in_use, add entrance-edge corners before computing stair_fit. Acceptance: ca/cb ratio ≈ 1.0 (currently 1.33).","status":"closed","priority":1,"issue_type":"task","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-13T18:08:11Z","created_by":"Bruno Postle","updated_at":"2026-06-13T19:53:40Z","started_at":"2026-06-13T19:09:35Z","closed_at":"2026-06-13T19:53:40Z","close_reason":"Fixed: entrance corner logic via _entrance_bid_for_stair (mirrors Perl Entrances), plus root cause: _avg_path_len_from now uses weighted Dijkstra (centroid distances) matching Perl graph.average_path_length — fixes has_circulation edge removal order. All 4 debug prefixes ratio=1.000, 39 tests pass.","dependencies":[{"issue_id":"homemaker-py-w1e","depends_on_id":"homemaker-py-hgg","type":"blocks","created_at":"2026-06-13T19:08:29Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":1,"comment_count":0}
 {"id":"homemaker-py-q70","title":"Fix corners_in_use _ib(None,None) bug: triple-at-idx=3 always passes in Perl","description":"In graph.py corners_in_use(), the _ib helper returns False when pa=None or pb=None. Perl's is_between_2d(point, undef, undef) returns True (distance_2d(undef,x)=0 so abs(0-0-0)\u003c1e-6). At triple-check idx=3, c1=corners[4]=None and c2=None, so _ib(w, None, None) must return True to match Perl — meaning the triple always succeeds at idx=3. Fix: add 'if pa is None and pb is None: return True' before the existing 'if pa is None or pb is None: return False'. This is already applied to graph.py. Needs: run 35-file corpus parity test to confirm aa0dcab98927d2c9 passes (corners [0,1,3] → stair_fit=0.878 → sf_factor≈0.570 ≈ oracle).","status":"closed","priority":1,"issue_type":"bug","owner":"bruno@postle.net","created_at":"2026-06-13T18:07:54Z","created_by":"Bruno Postle","updated_at":"2026-06-13T19:53:53Z","closed_at":"2026-06-13T19:53:53Z","close_reason":"Fixed as part of homemaker-py-w1e: the _avg_path_len_from weighted Dijkstra fix corrects has_circulation edge removal ordering, which was the actual cause of wrong stack corner counts. The _ib(None,None)=True fix was already in place but the weighted path length was the remaining blocker.","dependencies":[{"issue_id":"homemaker-py-q70","depends_on_id":"homemaker-py-hgg","type":"blocks","created_at":"2026-06-13T19:08:28Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":1,"comment_count":0}
@@ -10,7 +11,7 @@
 {"id":"homemaker-py-8cs","title":"Experiment: warm-vs-cold start of inner loop (Lamarckian inheritance)","description":"DESIGN.md §5.6, §4.6. Warm-starting a child topology's inner loop from the parent's optimised ratios is the main lever for cutting per-topology cost (~3 min/topology cold). Apply single topology mutations to optimised corpus designs, re-optimise warm (surviving cuts keep values, new cuts get heuristic defaults) vs cold, compare oracle-call counts to convergence at equal final fitness.","acceptance_criteria":"Speedup factor measured across \u003e=10 mutated topologies; decision recorded (expect order-of-magnitude; if \u003c2x, revisit §4.6 Phase-2 scoping)","notes":"Experiment script committed (experiments/warm_vs_cold.py, 1cc86c8) and machinery validated oracle-free; one mutated child scored through the oracle OK. Waiting on homemaker-py-gp2 reference run to finish, then execute under URB_NO_OCCLUSION=1 (3 parents x 400 evals + 12 children x 2 x 200 evals, ~1.5-2 h oracle time). Default budgets: parent 400, child 200; target = evals to 95% of best final.","status":"closed","priority":1,"issue_type":"task","owner":"bruno@postle.net","created_at":"2026-06-11T23:36:58Z","created_by":"Bruno Postle","updated_at":"2026-06-12T11:44:45Z","closed_at":"2026-06-12T11:44:45Z","close_reason":"Measured (URB_NO_OCCLUSION=1, parent budget 400, child 200, 12 single mutations across 3 designs): cold start reached 95% of warm final in 0/12 cases within budget — speedup unbounded at practical budgets; warm finals beat cold finals x1.2-x4 in 12/12; 6/12 warm starts were within 95% at 1 eval (near-neutral mutations). Decision: Lamarckian warm-starting is MANDATORY in the memetic driver (homemaker-py-b39), not an optimisation; cold starts produce strictly worse geometry at equal budget. Note: 2 undivides were exactly fitness-neutral (same-type merge == Merge_Divided equivalence) — locality datum for homemaker-py-nyb.","dependencies":[{"issue_id":"homemaker-py-8cs","depends_on_id":"homemaker-py-1p0","type":"blocks","created_at":"2026-06-12T00:39:34Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-av5","title":"Batched oracle: score many .dom files per invocation","description":"oracle.py currently scores one .dom per urb-fitness.pl call (~1.65 s/dom). DESIGN.md §4.6: batching amortises Perl startup to ~0.99 s/dom and is required so population/batch optimisers can score a whole generation in one oracle call. Extend oracle.py with a batch API: write N .dom files, one perl invocation, parse N .score/.fails pairs. Keep the single-file path for compatibility.","acceptance_criteria":"Batch of 35 corpus files scores in one perl invocation; per-file results identical to single-file calls; measured s/dom reported","status":"closed","priority":1,"issue_type":"task","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-11T23:36:56Z","created_by":"Bruno Postle","updated_at":"2026-06-12T00:14:06Z","started_at":"2026-06-11T23:50:40Z","closed_at":"2026-06-12T00:14:06Z","close_reason":"score_batch() lands in oracle.py; 35-file corpus parity verified single-vs-batch (1e-12 rel fitness, exact fail sets); 0.98 s/dom batched vs 1.27 single, x1.30","dependency_count":0,"dependent_count":1,"comment_count":0}
 {"id":"homemaker-py-g0b","title":"homemaker-fitness: native Python CLI replacement for urb-fitness.pl","description":"We need a Python CLI tool that replicates the behaviour of urb-fitness.pl so we can score .dom files without shelling out to Perl. The tool should: accept .dom file paths as arguments (or glob *.dom in cwd if none given), load patterns.config and costs.config from cwd and parent dir (local overrides project-level), skip scoring if .score and .fails files are already newer than the .dom (unless FORCE_UPDATE env var is set), score each .dom using fitness.Fitness.score_with_fails(), write the score to \u003cdom\u003e.score (40-digit float format), write the failures to \u003cdom\u003e.fails, print the score to stderr. Expose as homemaker-fitness entry point in pyproject.toml and as python -m homemaker_layout.fitness_cmd module. This replaces the oracle.py shelling-out path for Phase 3 native fitness.","status":"closed","priority":2,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-14T12:32:29Z","created_by":"Bruno Postle","updated_at":"2026-06-14T16:17:21Z","started_at":"2026-06-14T12:32:52Z","closed_at":"2026-06-14T16:17:21Z","close_reason":"Implemented as homemaker_layout/fitness_cmd.py with homemaker-fitness entry point; exact score parity verified against urb-fitness.pl on corpus","dependency_count":0,"dependent_count":0,"comment_count":0}
-{"id":"homemaker-py-gpx","title":"Native fitness parity gap on multi-storey designs (~3.7%)","description":"During programme-house cold-start runs with the fixed level_add operator, the generated 2-storey design showed native=1.2388e-04 vs oracle=1.1944e-04 (3.7% gap), exceeding the 0.01% rel_tol in test_native_fitness_score_parity. All existing single-storey corpus files pass parity fine (73/73). Hypothesis: a subtle discrepancy in value or cost computation for multi-level trees — candidates are staircase quality, circulation connectivity, or per-storey cost accumulation. To investigate: score a sweep of known multi-storey corpus files natively vs oracle and identify which term diverges.","status":"open","priority":2,"issue_type":"bug","owner":"bruno@postle.net","created_at":"2026-06-14T09:35:34Z","created_by":"Bruno Postle","updated_at":"2026-06-14T09:35:34Z","dependency_count":0,"dependent_count":0,"comment_count":0}
+{"id":"homemaker-py-gpx","title":"Native fitness parity gap on multi-storey designs (~3.7%)","description":"During programme-house cold-start runs with the fixed level_add operator, the generated 2-storey design showed native=1.2388e-04 vs oracle=1.1944e-04 (3.7% gap), exceeding the 0.01% rel_tol in test_native_fitness_score_parity. All existing single-storey corpus files pass parity fine (73/73). Hypothesis: a subtle discrepancy in value or cost computation for multi-level trees — candidates are staircase quality, circulation connectivity, or per-storey cost accumulation. To investigate: score a sweep of known multi-storey corpus files natively vs oracle and identify which term diverges.","status":"closed","priority":2,"issue_type":"bug","owner":"bruno@postle.net","created_at":"2026-06-14T09:35:34Z","created_by":"Bruno Postle","updated_at":"2026-06-17T17:39:25Z","closed_at":"2026-06-17T17:39:25Z","close_reason":"Closed","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-hqw","title":"Make homemaker-py standalone: remove dependency on Perl Urb package","description":"Currently tests and fitness scoring depend on the Perl Urb package (urb-fitness.pl) and corpus files in /home/bruno/src/urb/examples/. The tool should be fully standalone and not require any external Perl packages or local urb corpus paths. This includes: bundling or reimplementing any needed reference data, making the native Python fitness the default path, and ensuring tests pass without /home/bruno/src/urb present.","status":"closed","priority":2,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-13T22:27:54Z","created_by":"Bruno Postle","updated_at":"2026-06-13T22:39:28Z","started_at":"2026-06-13T22:34:20Z","closed_at":"2026-06-13T22:39:28Z","close_reason":"Closed","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-0px","title":"Blank-slate cold-start initialisation","description":"The outer search stalls when starting from init.dom (Phase 2 gate: 18 fails after 2000 evals vs urb-evolve's 6). The root cause is single-seed topology mutation chaining — building structure one room at a time gives no gradient across the large zero-feasibility region. Fix requires multi-start bootstrap: generate a diverse initial population by random topology sampling, or a greedy room-placement initialiser that satisfies adjacency/level constraints before handing off to the memetic loop. Without this the tool is only useful for refining existing designs, not designing new buildings from scratch.","acceptance_criteria":"Cold-start from init.dom reaches comparable fail count to urb-evolve within equal eval budget; tested on programme-house","status":"closed","priority":2,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-13T21:52:15Z","created_by":"Bruno Postle","updated_at":"2026-06-13T22:28:58Z","started_at":"2026-06-13T22:24:02Z","closed_at":"2026-06-13T22:28:58Z","close_reason":"Bootstrap implemented: auto-detect bare-plot seed, generate pop_size random topologies, evaluate each at child_budget before memetic loop; 3 new tests all green","dependency_count":0,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-ccw","title":"Scaled topology search on native fitness","description":"DESIGN.md §7 Phase 3 closing step. Once native fitness passes corpus parity, re-run the Phase-2 memetic search at real scale (population/generations comparable to urb-evolve) on the native objective. This is the first point where the §1 scaling question gets a real answer.","acceptance_criteria":"Full-scale run on programme-house beats both urb-evolve and the small-scale Phase-2 result; larger programme attempted","status":"closed","priority":2,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-11T23:38:59Z","created_by":"Bruno Postle","updated_at":"2026-06-13T21:11:13Z","started_at":"2026-06-13T20:49:27Z","closed_at":"2026-06-13T21:11:13Z","close_reason":"Closed","dependencies":[{"issue_id":"homemaker-py-ccw","depends_on_id":"homemaker-py-uxz","type":"blocks","created_at":"2026-06-12T00:39:44Z","created_by":"Bruno Postle","metadata":"{}"},{"issue_id":"homemaker-py-ccw","depends_on_id":"homemaker-py-way","type":"blocks","created_at":"2026-06-12T00:39:45Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":2,"dependent_count":1,"comment_count":0}
@@ -30,16 +31,16 @@
 {"id":"homemaker-py-yg5","title":"Penalty reshaping: replace 0.5^n while preserving inner-loop protection","description":"DESIGN.md §4.7, §5.4, §7 Phase 4, §8.5. The 0.5^n cliff gives the outer search no gradient and rewards flag-count over geometry, but it also PROTECTS the inner loop from trading into new failures (§4.5). One fitness shape cannot naively be both soft outside and cliff-protected inside. Candidates: cliff-inside-inner-loop only, lexicographic (failure count first, score second), additive/soft, multi-objective Pareto. Must preserve the missing-space failure hierarchy (worse to drop a room than to have a poor one). Measure landscape + search outcomes; this helps Urb today too.","acceptance_criteria":"Chosen scheme documented with measurements: search improves while inner loop still never trades into new failures","status":"closed","priority":3,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-11T23:39:00Z","created_by":"Bruno Postle","updated_at":"2026-06-14T08:16:14Z","started_at":"2026-06-14T07:55:32Z","closed_at":"2026-06-14T08:16:14Z","close_reason":"Implemented lexicographic outer-search comparison (-n_fails, fitness). Inner loop unchanged (0.5^n cliff protection preserved). Experiment penalty_reshape.py confirms 0/9 fail regressions in inner loop and shows lex avoids the 3-fail trap that scalar hits 1/3 of the time. Fixed stale _CHILD_INNER_KW sigmas entry.","dependencies":[{"issue_id":"homemaker-py-yg5","depends_on_id":"homemaker-py-uxz","type":"blocks","created_at":"2026-06-12T00:39:46Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-9gp","title":"Canonical slicing encoding (normalized Polish expression) + shape feasibility","description":"DESIGN.md §5.5, §7 Phase 5. Representation upgrade once core lands: normalized Polish expression / skewed slicing tree (Wong–Liu) for redundancy-free, high-locality topology moves (M1/M2/M3); bottom-up shape-feasibility checks to prune infeasible topologies before the inner loop. Goal: scale to larger programmes. Excluded representations stay excluded (§2): no sequence-pair/B*-tree (non-slicing).","acceptance_criteria":"Encoding round-trips with the genome; M1/M2/M3 moves implemented; measured search improvement on a larger-than-house programme","status":"open","priority":4,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-06-11T23:39:02Z","created_by":"Bruno Postle","updated_at":"2026-06-11T23:39:02Z","dependencies":[{"issue_id":"homemaker-py-9gp","depends_on_id":"homemaker-py-ccw","type":"blocks","created_at":"2026-06-12T00:39:48Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
 {"id":"homemaker-py-2g5","title":"Rebuild occlusion/daylight/sun subsystem in Python (post-Phase-5, after optimisation fully native)","description":"DESIGN.md §6 port scope — a whole subsystem, not a term. quality_daylight (Leaf.pm:281-296) needs Urb::Misc::Sun + Urb::Field::Occlusion (+CIESky); quality_uncrinkliness also takes the occlusion object. Indoor spaces return 1 for daylight; cost is outdoor spaces + crinkliness. Port Sun_horizontal (262980-minute normalisation) and the occlusion wall set from Dom-\u003eWalls.","acceptance_criteria":"Daylight and crinkliness factors match Perl (float tolerance) across the corpus, including multi-storey cases","notes":"Re-scoped 2026-06-12: occlusion disabled in the Urb oracle instead of ported (see homemaker-py-gp2). Native fitness ships with simple crinkliness (illumination factor = 1, in homemaker-py-gnw). This issue is now the eventual Python occlusion rebuild, only after optimisation works entirely in Python. Restores outdoor-daylight and shaded-wall selection pressure.","status":"open","priority":4,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-06-11T23:38:25Z","created_by":"Bruno Postle","updated_at":"2026-06-12T07:27:48Z","dependency_count":0,"dependent_count":0,"comment_count":0}
-{"_type":"memory","key":"adjacency-in-binary-slicing-tree-is-structural-not","value":"Adjacency in binary slicing tree is structural, not geometric: the inner-loop NM cannot fix topological adjacency failures. Two paths exist: (1) tree-sibling adjacency — a node is adjacent to its sibling in the tree; (2) cross-zone geometric adjacency — leaves from different subtrees that happen to share a boundary. Staircase/adjacency fails require a topology mutation that changes which nodes are siblings or which zones touch. This was proved empirically on programme-house: staircase fail from rot=0 layout could not be fixed by NM but was fixed by level_retype creating a two-C topology (2026-06-14/15)."}
-{"_type":"memory","key":"cli-tool-style-prefer-python-m-homemaker-module","value":"CLI tool style: prefer python -m homemaker.module --parameters pattern, installable via pip install -e . with pyproject.toml entry_points. Not standalone bin/ scripts."}
-{"_type":"memory","key":"correction-to-urb-fitness-bug-memory-bruno-2026","value":"CORRECTION to urb-fitness-bug memory (Bruno, 2026-06-12): 'C' is NOT a 'covered' type — Is_Covered is a geometric predicate (indoor space above). Urb's generic types are canonically UPPERCASE: C=circulation, O=outside, S=sahn (get_space_types qw/C O S/; corpus is 100% uppercase, never 'c'/'o' leaves). The mixed-case designs that fired the latent ratio_type first-match bug were created by homemaker's own operator type pool emitting lowercase 'c'/'o' — fixed: driver/operators now emit uppercase generics only, and class checks use t[0].lower() in 'cos'. The Urb class-sum patch stays as defensive hardening (zero impact on canonical designs). Native port (3y7/gnw): treat type classes case-insensitively, generics canonically uppercase."}
-{"_type":"memory","key":"deceptive-valleys-in-topology-search-when-every-single","value":"Deceptive valleys in topology search: when every single-step mutation from a target state passes through a high-fail intermediary (e.g. level_fix displaces a room into 5+ new fails), a compound operator that atomically applies two coordinated changes can escape. Design compound operators to land on the low-fail state directly, bypassing the deceptive gradient. Programme-house example: level_compound_fix atomically moves the level-constrained room AND re-inserts the displaced room adjacent to C in one step (operators.py, 2026-06-14)."}
+{"_type":"memory","key":"multi-storey-staircase-consistency-when-dividing-or-retyping","value":"Multi-storey staircase consistency: when dividing or retyping a circulation (C) leaf at one level, the same structural change should be propagated to the matching leaf on ALL other storeys so the stair core path is maintained. The optimizer cannot fix staircase disruptions through trial-and-error geometry alone — it requires a synchronized multi-level operator that applies the same topology change to every storey simultaneously."}
+{"_type":"memory","key":"strategy-decision-2026-06-12-bruno-occlusion-daylight","value":"Strategy decision 2026-06-12 (Bruno): occlusion/daylight is ORTHOGONAL to building a scalable optimiser. Disable it in Urb (env flag, homemaker-py-gp2) rather than port it; native fitness uses simple crinkliness (illumination factor = 1); rebuild occlusion in Python only after optimisation is fully native (homemaker-py-2g5, now P4). Consequence: all scores change when the flag flips — re-baseline corpus/.score, DESIGN \\$4.5 gains, gate bars at one clean boundary AFTER homemaker-py-1p0 closes; Phase-2 urb-evolve benchmark must run with the same flag."}
+{"_type":"memory","key":"programme-house-optimisation-result-2026-06-14-15","value":"Programme-house optimisation result (2026-06-14/15): best achievable is 1 fail (l1 wrong level, score ~0.005). 0 fails is geometrically impossible: l1 (min 27m²) must occupy ll (~23m²) at level 0, which eliminates the t3-adj-C provider; dividing ll into lll(l1)+llr(C) gives llr proportion ~6:1 (fails). Python memetic optimizer achieves 1 fail in 50k evals vs Perl optimiser's 2-3 fails. Winning topology: TWO C nodes at level 0 — ll(C) for t3-adj-C via geometric contact, rl(C) for staircase via tree-sibling adjacency to rrr(O). Best .dom: scratch/from-warmstart-fixed.dom and scratch/from-compound3-fixed.dom."}
 {"_type":"memory","key":"urb-fitness-bug-found-fixed-2026-06-12","value":"Urb fitness bug found+fixed 2026-06-12 (patch in /home/bruno/src/urb, uncommitted): ProgrammeDriven.pm ratio_o/ratio_type grepped case-insensitively over the ratios hash and took the FIRST key — nondeterministic (x4.5 score swings) for designs with mixed-case type classes (both 'c' circulation and 'C' covered). Fixed to SUM the class (matches Is_Circulation//Is_Outside semantics); 35/35 corpus scores unchanged. CRITICAL for homemaker-py-3y7/gnw: the native port must implement class-SUM ratios. Building.pm has the same unpatched pattern (site-driven path, not used by our oracle). Also: the memetic search reward-hacked this bug before the fix — search results predating it are noise artifacts."}
 {"_type":"memory","key":"user-preference-bruno-this-is-a-fedora-system","value":"User preference (Bruno): this is a Fedora system — NEVER install Python packages via pip without asking first; always ask whether to install the rpm via dnf (e.g. python3-cma) before considering pip. Applies to any dependency additions."}
 {"_type":"memory","key":"warm-x0-initialization-bug-pattern-when-a-topology","value":"warm_x0 initialization bug pattern: when a topology operator explicitly sets division ratios on a newly-created node (e.g. compound_fix sets node.division=[0.25,0.25] for t3), parent.ratios has no entry for that node (it was a leaf). warm_x0 defaults it to 0.5, corrupting the inner loop's starting point and making the operator invisible to lex comparison. Fix: only propagate child ratios for nodes where the parent node was NOT already divided; stale hidden nodes revealed by structural mutations (swap flipping b.below) must NOT contribute their pre-writeback values. See driver.py lines 259-267 (fixed 2026-06-14)."}
-{"_type":"memory","key":"never-use-corpus-filenames-candidate-001-dom-candidate","value":"Never use corpus filenames (candidate-001.dom, candidate-002.dom, generated.dom, init.dom, etc.) as --output targets when running experiments. These are test fixtures. Always write experimental outputs to scratch/ or a timestamped path. Lesson from 2026-06-14: warm-start runs overwrote candidate-001/002.dom and broke graph tests."}
-{"_type":"memory","key":"programme-house-optimisation-result-2026-06-14-15","value":"Programme-house optimisation result (2026-06-14/15): best achievable is 1 fail (l1 wrong level, score ~0.005). 0 fails is geometrically impossible: l1 (min 27m²) must occupy ll (~23m²) at level 0, which eliminates the t3-adj-C provider; dividing ll into lll(l1)+llr(C) gives llr proportion ~6:1 (fails). Python memetic optimizer achieves 1 fail in 50k evals vs Perl optimiser's 2-3 fails. Winning topology: TWO C nodes at level 0 — ll(C) for t3-adj-C via geometric contact, rl(C) for staircase via tree-sibling adjacency to rrr(O). Best .dom: scratch/from-warmstart-fixed.dom and scratch/from-compound3-fixed.dom."}
+{"_type":"memory","key":"correction-to-urb-fitness-bug-memory-bruno-2026","value":"CORRECTION to urb-fitness-bug memory (Bruno, 2026-06-12): 'C' is NOT a 'covered' type — Is_Covered is a geometric predicate (indoor space above). Urb's generic types are canonically UPPERCASE: C=circulation, O=outside, S=sahn (get_space_types qw/C O S/; corpus is 100% uppercase, never 'c'/'o' leaves). The mixed-case designs that fired the latent ratio_type first-match bug were created by homemaker's own operator type pool emitting lowercase 'c'/'o' — fixed: driver/operators now emit uppercase generics only, and class checks use t[0].lower() in 'cos'. The Urb class-sum patch stays as defensive hardening (zero impact on canonical designs). Native port (3y7/gnw): treat type classes case-insensitively, generics canonically uppercase."}
+{"_type":"memory","key":"deceptive-valleys-in-topology-search-when-every-single","value":"Deceptive valleys in topology search: when every single-step mutation from a target state passes through a high-fail intermediary (e.g. level_fix displaces a room into 5+ new fails), a compound operator that atomically applies two coordinated changes can escape. Design compound operators to land on the low-fail state directly, bypassing the deceptive gradient. Programme-house example: level_compound_fix atomically moves the level-constrained room AND re-inserts the displaced room adjacent to C in one step (operators.py, 2026-06-14)."}
 {"_type":"memory","key":"homemaker-py-pythonpath-set-pythonpath-home-bruno-src","value":"homemaker-layout PYTHONPATH: package installed as 'homemaker-layout' via pip install -e . so 'import homemaker_layout' works from anywhere without PYTHONPATH. For running tests use 'python -m pytest' from project root /home/bruno/src/homemaker-layout (pyproject.toml adds src/ automatically). Never try pip show homemaker — that's the old homemaker-addon conflict."}
-{"_type":"memory","key":"multi-storey-staircase-consistency-when-dividing-or-retyping","value":"Multi-storey staircase consistency: when dividing or retyping a circulation (C) leaf at one level, the same structural change should be propagated to the matching leaf on ALL other storeys so the stair core path is maintained. The optimizer cannot fix staircase disruptions through trial-and-error geometry alone — it requires a synchronized multi-level operator that applies the same topology change to every storey simultaneously."}
-{"_type":"memory","key":"strategy-decision-2026-06-12-bruno-occlusion-daylight","value":"Strategy decision 2026-06-12 (Bruno): occlusion/daylight is ORTHOGONAL to building a scalable optimiser. Disable it in Urb (env flag, homemaker-py-gp2) rather than port it; native fitness uses simple crinkliness (illumination factor = 1); rebuild occlusion in Python only after optimisation is fully native (homemaker-py-2g5, now P4). Consequence: all scores change when the flag flips — re-baseline corpus/.score, DESIGN \\$4.5 gains, gate bars at one clean boundary AFTER homemaker-py-1p0 closes; Phase-2 urb-evolve benchmark must run with the same flag."}
 {"_type":"memory","key":"urb-oracle-nondeterminism-urb-fitness-pl-output-varies","value":"Urb oracle nondeterminism: urb-fitness.pl output varies run-to-run from Perl hash-order randomisation — .fails line ORDER shuffles (compare sorted, use oracle.Score.fail_lines) and the score float can flip by ~1 ULP (compare with math.isclose rel_tol=1e-12, never ==). Not a batching artifact; affects single runs too. Matters for the Phase 3 native-fitness parity gate (homemaker-py-uxz)."}
+{"_type":"memory","key":"adjacency-in-binary-slicing-tree-is-structural-not","value":"Adjacency in binary slicing tree is structural, not geometric: the inner-loop NM cannot fix topological adjacency failures. Two paths exist: (1) tree-sibling adjacency — a node is adjacent to its sibling in the tree; (2) cross-zone geometric adjacency — leaves from different subtrees that happen to share a boundary. Staircase/adjacency fails require a topology mutation that changes which nodes are siblings or which zones touch. This was proved empirically on programme-house: staircase fail from rot=0 layout could not be fixed by NM but was fixed by level_retype creating a two-C topology (2026-06-14/15)."}
+{"_type":"memory","key":"cli-tool-style-prefer-python-m-homemaker-module","value":"CLI tool style: prefer python -m homemaker.module --parameters pattern, installable via pip install -e . with pyproject.toml entry_points. Not standalone bin/ scripts."}
+{"_type":"memory","key":"never-use-corpus-filenames-candidate-001-dom-candidate","value":"Never use corpus filenames (candidate-001.dom, candidate-002.dom, generated.dom, init.dom, etc.) as --output targets when running experiments. These are test fixtures. Always write experimental outputs to scratch/ or a timestamped path. Lesson from 2026-06-14: warm-start runs overwrote candidate-001/002.dom and broke graph tests."}
diff --git a/CLAUDE.md b/CLAUDE.md
index 082fa79..f30b452 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -52,27 +52,49 @@ bd close <id>         # Complete work
 
 ## Build & Test
 
-_Add your build and test commands here_
-
 ```bash
-# Example:
-# npm install
-# npm test
+pip install -e .
+pytest
 ```
 
 ## Architecture Overview
 
-_Add a brief overview of your project architecture_
+homemaker-layout is a Python successor to the Perl [Urb](../urb) project. It
+represents a building as a binary slicing tree where leaves carry **target
+dimensions** from the programme and division ratios are **solved bottom-up**
+(inverting Urb's top-down approach). The evolutionary search explores topology,
+types, and adjacency only.
+
+Key modules:
+- `dom.py` — read/write Urb `.dom` YAML into a `Node` tree
+- `geometry.py` — faithful port of Urb's top-down geometry
+- `programme.py` — parse `patterns.config` space requirements
+- `solver.py` — bottom-up ratio solve (scipy)
+- `fitness.py` — native Python fitness evaluator (replaces Perl oracle)
+- `fitness_cmd.py` — `homemaker-fitness` CLI entry point
+- `graph.py` — leaf-adjacency graph for programme-driven fitness checks
+- `genome.py` — topology genome: base-floor tree + per-storey deltas
+- `operators.py` — high-locality mutation and subtree crossover
+- `innerloop.py` — ratio optimisation inner loop (Nelder-Mead / CMA-ES)
+- `driver.py` — memetic search outer loop
+- `evolve.py` — `homemaker-evolve` CLI entry point
+- `oracle.py` — legacy Perl shim, kept for validation only; do not use in new code
 
 ## Conventions & Patterns
 
-### Oracle (urb-fitness.pl)
+### Scoring .dom files
 
-`urb-fitness.pl` is in PATH. To score a `.dom` file you **must `cd` to the directory containing the `.dom` file first** — the script resolves `patterns.config`, `costs.config`, and writes `.score`/`.fails` relative to `cwd`:
+Use the native `homemaker-fitness` command. Like the old `urb-fitness.pl`, you
+**must `cd` to the directory containing the `.dom` file first** — the tool
+resolves `patterns.config`, `costs.config`, and writes `.score`/`.fails`
+relative to `cwd`:
 
 ```bash
-cd /home/bruno/src/urb/examples/programme-house
-urb-fitness.pl cf0b8a77e8b2325f92a7e7d150184a55.dom
+cd /home/bruno/src/homemaker-layout/examples/programme-house
+homemaker-fitness cf0b8a77e8b2325f92a7e7d150184a55.dom
 ```
 
 The score is written to `<file>.dom.score` and failures to `<file>.dom.fails`; the numeric score is also printed to stderr.
+
+Do **not** use `urb-fitness.pl` directly — `oracle.py` and the Perl tool are
+kept only for cross-validation.
diff --git a/README.md b/README.md
index 3af2f13..d45bcbe 100644
--- a/README.md
+++ b/README.md
@@ -20,10 +20,11 @@ search then only explores topology + types + adjacency.
 
 ## Phase plan
 
-1. **Solver experiment** (current): port Urb's geometry, re-solve ratios from
-   programme targets, score the result against the original via the Perl oracle.
-2. Native Python fitness (retire the Perl oracle).
-3. Canonical slicing encoding (normalized Polish expression) + memetic search.
+1. ~~Solver experiment: port Urb's geometry, re-solve ratios from programme
+   targets, score the result against the original via the Perl oracle.~~ ✓
+2. ~~Native Python fitness (retire the Perl oracle).~~ ✓
+3. **Memetic search** (current): canonical slicing genome + high-locality
+   operators + Nelder-Mead inner loop.
 
 ## Layout
 
@@ -31,6 +32,12 @@ search then only explores topology + types + adjacency.
 - `src/homemaker_layout/geometry.py` — faithful port of Urb's top-down geometry.
 - `src/homemaker_layout/programme.py` — parse `patterns.config` space requirements.
 - `src/homemaker_layout/solver.py` — bottom-up ratio solve (scipy).
-- `src/homemaker_layout/oracle.py` — Phase-1 scaffold: score a `.dom` via Urb's `urb-fitness.pl`.
-
-The Perl oracle is the only throwaway component; everything else is permanent.
+- `src/homemaker_layout/fitness.py` — native Python fitness evaluator.
+- `src/homemaker_layout/fitness_cmd.py` — `homemaker-fitness` CLI (drop-in for `urb-fitness.pl`).
+- `src/homemaker_layout/graph.py` — leaf-adjacency graph for programme-driven checks.
+- `src/homemaker_layout/genome.py` — topology genome: base-floor tree + per-storey deltas.
+- `src/homemaker_layout/operators.py` — high-locality mutation and subtree crossover.
+- `src/homemaker_layout/innerloop.py` — ratio optimisation inner loop (Nelder-Mead / CMA-ES).
+- `src/homemaker_layout/driver.py` — memetic search outer loop.
+- `src/homemaker_layout/evolve.py` — `homemaker-evolve` CLI entry point.
+- `src/homemaker_layout/oracle.py` — legacy Perl shim, kept for cross-validation only.
diff --git a/src/homemaker_layout/oracle.py b/src/homemaker_layout/oracle.py
index 8f86e2d..fa7d91e 100644
--- a/src/homemaker_layout/oracle.py
+++ b/src/homemaker_layout/oracle.py
@@ -27,9 +27,101 @@ from dataclasses import dataclass
 from pathlib import Path
 from typing import Sequence
 
+import yaml
+
 DEFAULT_URB_ROOT = Path("/home/bruno/src/urb")
 
 
+def _structured_fail_to_str(f: dict) -> str:
+    """Convert a structured failure dict (llm-agent-mcp branch format) to the
+    plain-text string that master urb/ProgrammeDriven.pm would have emitted."""
+    t = f.get("type", "")
+    if t == "level":
+        return f"{f['space']} on wrong level (level {f['actual']}, expected {f['required']})"
+    if t == "missing":
+        code = f["code"]
+        c = f.get("constraint")
+        if c == "size":
+            return f"missing {code}: would need size check"
+        if c == "width":
+            return f"missing {code}: would need width check"
+        if c == "proportion":
+            return f"missing {code}: would need proportion check"
+        if c == "adjacency":
+            return f"missing {code}: would need adjacency to {f.get('target', '')}"
+        if c == "level":
+            return f"missing {code}: would need to be on level {f['required']}"
+        if c == "vertical":
+            return f"missing {code}: would need connection to {f.get('target', '')} below"
+        if f.get("critical"):
+            return f"missing required space: {code} (critical)"
+        return f"missing required space: {code}"
+    if t == "count":
+        return f"too many spaces: {f['code']} (found {f['actual']}, expected {f['expected']})"
+    if t == "adjacency":
+        return f"{f['node']} ({f['space']}) not adjacent to {f['target']}"
+    if t == "vertical":
+        return f"{f['space']} not connected to {f['target']} below"
+    if t == "staircase":
+        issue = f.get("issue", "")
+        if issue == "volume":
+            return "staircase volume"
+        if issue == "count":
+            actual = f.get("actual", 0)
+            if "min" in f:
+                return f"too few stairs ({actual}, min {f['min']})"
+            if "max" in f:
+                return f"too many stairs ({actual}, max {f['max']})"
+    if t == "storey":
+        if f.get("issue") == "limit":
+            return "storey limit"
+        if f.get("issue") == "minimum":
+            return "storey minimum"
+    if t == "access" and f.get("issue") == "no_outside_public_access":
+        return "no outside public access"
+    return str(f)
+
+
+def _parse_fails(text: str) -> list[str]:
+    """Parse a .fails file that may contain a YAML block followed by plain-text
+    lines (urb branch format) or only plain-text lines (master format)."""
+    text = text.strip()
+    if not text:
+        return []
+    if not text.startswith("---"):
+        return [line.strip() for line in text.splitlines() if line.strip()]
+
+    # Split YAML block from trailing plain-text lines: YAML list items start
+    # with "- " or are indented; once we hit a non-indented, non-dash line
+    # that isn't blank or the "---" marker, the YAML part has ended.
+    yaml_lines: list[str] = []
+    plain_lines: list[str] = []
+    in_yaml = True
+    for line in text.splitlines():
+        if in_yaml:
+            stripped = line.strip()
+            if not stripped or stripped == "---" or stripped.startswith("-") or line[:1] == " ":
+                yaml_lines.append(line)
+            else:
+                in_yaml = False
+                plain_lines.append(stripped)
+        else:
+            if line.strip():
+                plain_lines.append(line.strip())
+
+    result: list[str] = []
+    try:
+        doc = yaml.safe_load("\n".join(yaml_lines))
+        if isinstance(doc, list):
+            for item in doc:
+                if isinstance(item, dict):
+                    result.append(_structured_fail_to_str(item))
+    except yaml.YAMLError:
+        pass
+    result.extend(plain_lines)
+    return result
+
+
 @dataclass
 class Score:
     fitness: float
@@ -40,9 +132,7 @@ class Score:
         """Failure messages as a sorted tuple — Perl's per-process hash-order
         randomisation shuffles the raw ``.fails`` line order between runs, so
         comparisons must be order-insensitive."""
-        return tuple(
-            sorted(line.strip() for line in self.fails.splitlines() if line.strip() and line.strip() != "---")
-        )
+        return tuple(sorted(_parse_fails(self.fails)))
 
     @property
     def n_fails(self) -> int:
@@ -68,7 +158,7 @@ def score_batch(
         Path(f"{p}.fails").unlink(missing_ok=True)
 
     urb_root = Path(urb_root).resolve()
-    env = {**os.environ, "DEBUG": "1"}
+    env = {**os.environ, "DEBUG": "1", "URB_NO_OCCLUSION": "1"}
     proc = subprocess.run(
         ["perl", f"-I{urb_root}/lib", str(urb_root / "bin" / "urb-fitness.pl")]
         + [p.name for p in paths],