{
  "title": "Otterchip Model E \u2014 independent manufacturing-implementation review: FINAL release disposition",
  "date": "2026-11-06",
  "revision": "final (bounded known-item disposition, reviewed at about 10:25\u201310:35Z). It supersedes the earlier interim revision, whose findings stay below as provenance.",
  "reviewer": "Claude Opus 5.5 (claude-opus-5-5), effort high; independent reviewer writing only hardware/otterchip_model_e/review/**. No mechanical, app or optimizer files were edited.",
  "scope": "Bounded check of the known items only: BLK-1 to BLK-3, D-1 to D-5, stale instrument_deck assembled exports, parameter identity and final scores. No new tests or research. Geometric collision values are taken from the mechanical QA. Artifact hashes and identity come from the native audit, not from me.",
  "evidence": {
    "current": {
      "mechanicalQA": {
        "file": "mechanical/out/qa/qa.json",
        "generated": "2026-11-06T05:24:51.505517Z",
        "revision": "E-mech-r1 (build 9)"
      },
      "partManifest": {
        "file": "mechanical/out/docs/part-manifest.json",
        "generated": "2026-11-06T05:24:53.202847Z"
      },
      "releaseStatus": {
        "file": "manufacturing/release-status.json",
        "generated": "2026-11-06T05:25:46.902747Z"
      },
      "assemblyGuide": "mechanical/out/docs/assembly-instructions.html (build 9 docs)",
      "nativeAudit": {
        "file": "review/native-manufacturing-audit.json",
        "generated": "2026-11-06T10:25:02.417950Z",
        "checks": 746,
        "failures": 0,
        "qaGenerated": "2026-11-06T05:24:51.505517Z",
        "manifestGenerated": "2026-11-06T05:24:53.202847Z",
        "note": "Native work: hashes, identity, topology metadata and staging consistency. I cite it and do not claim it."
      },
      "derivedRegression": "mechanical/derive-regression.json (00:25 local): ok true, 68 rows, 0 failing",
      "sectionImage": "manufacturing/model-e-sections.png (viewed; see sectionImageReview)",
      "optimizer": "optimization/results.json generatedAt 2026-11-06T05:03:35.700Z, core 1.1.0-model-e-frozen. optimization/native/selected-check.json gives J 0.014797756266410417, equal to results.json."
    },
    "staleSupervisorWrappers": {
      "files": [
        "mechanical/assembly-manifest.json (00:00 local)",
        "mechanical/native-export-audit.json (00:12 local)"
      ],
      "status": "Stale supervisor metadata predating build 9, awaiting root regeneration. They are NOT the current evidence; out/qa/qa.json and out/docs/part-manifest.json are. This is a metadata refresh only, not a reason for a CAD rebuild. Root gates packaging on refreshed hashes."
    },
    "provenanceSnapshots": [
      "QA 05:01:05Z: 2 service failures (wing R 21.12 mm\u00b3; keybed 5985.85 mm\u00b3).",
      "QA 05:04:04Z: service paths passed; shell mesh/BRep 2.86% from stale relative tessellation.",
      "QA 05:06:45Z: shell mismatch fixed by absolute tessellation."
    ]
  },
  "parameterIdentity": {
    "result": "verified",
    "detail": "QA selected_parameters equal the root-frozen width 355 / rear 92 / height 245 / tilt 18 / setback 180 / deck 62.5 / slope 0 / drop 23 / underlap 39 / pad 7/1/10. The QA stack gives key top 73.3 mm; the section image title gives depth 415.4 and height 338.0, matching the viewer (355 \u00d7 338.0 \u00d7 415.4)."
  },
  "currentAutomatedState_build9": {
    "records": 110,
    "printedParts": 58,
    "selectedFirstBuildPrinted": 36,
    "brepValidSingleSolidPrinted": true,
    "meshesWatertightWindingSingleComponent": true,
    "stepRoundTrip": true,
    "interferences": 0,
    "allowedDocumentedOverlaps": 4,
    "fastenerChecks": 87,
    "fastenerFailures": 0,
    "servicePaths": "9 of 9 pass (solid sweeps only; hands, tools, wires and part flex not modelled)",
    "maxMeshVolumeRelErr": "0.00701 (volume_roller_tyre_tpu). Next highest is 0.00272 (shim_0p5mm); the shells are accurate.",
    "printStlPerStage": {
      "A_fit_coupons": 17,
      "A_fit_and_beds": 4,
      "A_acoustic_prototype": 2,
      "B_prototype_parts": 31,
      "B_prototype_shell": 3,
      "B_prototype_shell_alt": 1
    },
    "assembledExports": "110 STEP + 110 STL. Obsolete instrument_deck.step/.stl are gone; only instrument_deck_rear and wing_L/R are present."
  },
  "dispositions": [
    {
      "id": "BLK-1",
      "was": "Shell print STL vs BRep volume mismatch (2.86%) from stale relative tessellation",
      "status": "RESOLVED",
      "evidence": "Build 9 QA: every printed mesh is watertight, winding-consistent and single-component, and every STEP round-trips. Shells agree closely with BRep after absolute tessellation (exportStl relative=False, build.py ~279\u2013299). The only remaining difference is the flexible volume-roller tyre: 2.473 cm\u00b3 BRep vs 2.490 cm\u00b3 mesh, about 17 mm\u00b3 (0.70%), on a \u00d8~50 \u00d7 8 mm TPU tread. Spread over that surface, 17 mm\u00b3 is of order 0.01 mm. That is negligible next to the designed 0.4 mm diametral stretch fit and TPU process tolerance, so it is accepted as a minor tessellation approximation of the flexible tread. My earlier '\u22640.5% for every part' threshold was my own heuristic, not a manufacturing standard, and is withdrawn as a gate."
    },
    {
      "id": "BLK-2",
      "was": "Docs reported the 28.55 mm ramp envelope as black-key exposure",
      "status": "RESOLVED",
      "evidence": "QA model_e_stack now separates black_key_flat_top_ahead_of_cad_bridge 25.05 (the ergonomic gate number) from black_key_outer_envelope_ahead_of_cad_bridge 28.55 and the viewer-edge value 26.1, with a note on the 1.05 mm CAD lip offset. The assembly guide states 'flat black-key top ahead of the CAD typing lip is 25.05 mm nominal (outer key envelope 28.55 mm; viewer keyboard edge gives 26.10 mm)' and requires a hand test. The optimizer's blackKeyProbe gives cadFingerClearanceMm 8.68 (fingertip radius estimated, CAD lip 1.05 mm forward), above the 3 mm allowance. The white-target bridge clearance of 40 mm in the broad pose is not proof for the whole playing area. All of this rests on donor photo estimates."
    },
    {
      "id": "BLK-3",
      "was": "Guide lacked the dock exclusion and the service/mesh status",
      "status": "RESOLVED",
      "evidence": "The assembly guide now says: 'rear cover-storage dock is a viewer concept with no CAD or BOM here'; 'Service paths: 9 of 9 passed as solid sweeps; hands, tools and wires are not modelled'; 'Mesh versus BRep volume: largest difference 0.70%, on the small flexible volume-roller tyre'; 'Do not order the large shell until gates G0-G5 and G9 close with measured values'; and 'No load rating: do not sit, stand or lean on any part'. release-status.json lists the dock and any sitting provision as excluded."
    },
    {
      "id": "D-1",
      "was": "parts.py:207 docstring said MR63ZZ",
      "status": "RESOLVED",
      "evidence": "parts.py:207 now reads 'MR84ZZ (4x8x3) bearings...', consistent with parameters, BOM and docs."
    },
    {
      "id": "D-2",
      "was": "derive-regression.json stale ok:false",
      "status": "RESOLVED",
      "evidence": "derive-regression.json (00:25 local): ok true, 68 rows, 0 failing."
    },
    {
      "id": "D-3",
      "was": "Backing-bed screw engagement depended on the shim",
      "status": "RESOLVED (documented)",
      "evidence": "The guide's F14 section gives floor 3.2 + land 1.3 = 4.5 mm before the shim, with at least 2 mm engagement and at least 0.3 mm tip recess. Table: 0.5 mm shim \u2192 trim an 8 mm screw to 7.5; 1.0 mm \u2192 8 mm as supplied; 2.0 mm \u2192 trim a 10 mm screw to 9. Each gives 2.5 mm engagement and 0.5 mm recess. The measured stack overrides the table, and no protruding screw is allowed."
    },
    {
      "id": "D-4",
      "was": "gasket_chamber_R was 2 solids vs L 1",
      "status": "RESOLVED (fixed)",
      "evidence": "Build 9 QA gives gasket_chamber_R solids 1. The guide states 'one closed piece per side, passing the CAD contact check. This is not a leak test.'"
    },
    {
      "id": "D-5",
      "was": "Shim quantity and 3MF collection semantics",
      "status": "RESOLVED (documented)",
      "evidence": "The guide states that each STL is one part geometry, quantity is not multiplied, and stage 3MF bundles mix PA12, TPU and conditional parts, so they are inspection collections, not print jobs. Quantities come from print-schedule.csv (e.g. 6 of each shim). Flexible parts are process-gated by G0."
    },
    {
      "id": "stale-exports",
      "was": "Obsolete instrument_deck.step/.stl",
      "status": "RESOLVED",
      "evidence": "110 assembled STEP and 110 STL; only instrument_deck_rear and instrument_deck_wing_L/R remain. The native audit (746 checks, 0 failures) checks for extra exports."
    }
  ],
  "acceptedSourceFixesRetained": [
    "Lowered deck: 3 mm aluminium backing bed on six 1.3 mm floor lands with fork shims and CSK M3 from below (parts.py:167\u2013204; shell.py:179\u2013203). It replaces Model M's inverted printed rails.",
    "Wheel cradle base opened around the full disc; wheel-to-floor gap 6.8 mm (parts.py:236\u2013238).",
    "Side controls: jacks on the left wall, rocker on the right wall, port plates withdraw inward.",
    "The keybed + backing bed + rear strip service unit lifts 3, moves forward 42, then lifts out. The bed screws must be removed from below first.",
    "The four allowed overlaps each have a stated design intent: mic TPU squeeze \u00d72, volume tyre stretch, Pi standoffs in the wire keep-out zone."
  ],
  "sectionImageReview": {
    "image": "hardware/otterchip_model_e/manufacturing/model-e-sections.png (viewed)",
    "observations": [
      "The title matches the final build: deck 62.5, drop 23.0, key top 73.3, depth 415.4, height 338.0, 'geometry only; donors are photo estimates; prototype CAD, not production approved'.",
      "Centre section (x = 0): low typing deck at ~62 mm, the keybed and backing bed close to the floor at the front, a single fluid sweep up to the 18\u00b0-tilted display, and head roof ~338. These match the web proportions.",
      "Left driver/keybed section (x = \u2212125): the speaker chamber sits in the head behind the display, and the low music stack is unchanged.",
      "Wheel/floor-bar section (x = \u2212156): the wheel circle sits just above the floor bar with visible clearance, and the cradle sits low. Consistent with the QA wheel-to-floor gap of 6.8 mm.",
      "Side-port section (viewer z = \u221280): left jack block and right rocker sit low inside the side walls near the table plane, inside the shell outline.",
      "Outlines that cross in a 2D section are planar projections of separate parts at one cut. They are not taken as collisions; collisions are judged from QA BRep interference (0). Outline thickness and shape are not taken as evidence of strength."
    ]
  },
  "readiness": {
    "exportGeometry": "Build 9 automated geometry and export checks pass: BRep and topology, mesh, STEP, interference, fasteners and service sweeps.",
    "stageA": "Stage A coupon and fit files (17 coupons, 4 fit/bed parts, 2 optional acoustic parts) can be prepared for bureau quotation and printing using the selected materials and the print-schedule quantities.",
    "stageB_parts": "Technically gated on coupon results (insert, tolerance and bearing ladders, jack window, keybed gauge, display seat) and on measured donors and display.",
    "fullShell": "Conditional. Order only after measured donor/display fit, coupon results and the acoustic prototype close the documented gates (guide: G0\u2013G5 and G9). Use the split 3-part shell; shell_onepiece is a large-frame alternative; the display spacer frame is conditional on the measured monitor.",
    "production": "NOT validated. No sit/lean load, tip, drop/impact, thermal, electrical/EMC, acoustic or child-use tests and no FEA have been performed. Passing export geometry does not establish any of these, and hollow PA12 outlines say nothing about strength."
  },
  "remainingTechnicalGates": [
    "Coupon measurements fed back into the parameters.",
    "Opened-donor measurements: LPK25 reserve, flat black-key datum, frame and screws, controller ribbon. Also the GK64X stack, AMOLED module (resolve display-source conflict G1), 3382H sensor and MR84ZZ bearings.",
    "Hand test on cpn_keybed_profile_gauge with a mock bridge (black keys).",
    "Acoustic prototype: sealing, buzz and driver motion.",
    "Physical validation: sit/lean loads, tip stability (E is taller and deeper than M; ballast bays untested), drop, small parts and pinch, thermal, electrical/EMC, child fit and playing trials."
  ],
  "packagingNote": "Root owns packaging, refreshed hashes and any manufacturer-reference redistribution. Not reviewed here.",
  "limitations": [
    "This was a bounded known-item check. I did not run the CAD build or new geometry tests.",
    "Collision and service values are the mechanical QA's.",
    "Donor geometry is photo-derived."
  ],
  "nativeReleaseWrapperReconciliation": {
    "reviewer": "Native release coordinator, separate from actual Claude review",
    "generatedUtc": "2026-11-06T10:28:35.496803+00:00",
    "qaGenerated": "2026-11-06T05:24:51.505517+00:00",
    "qaSHA256": "66c1f2dc42807a5d19fd7b1d60ee5fbb6cbf5b906fc4a96097717ff638bc08f9",
    "assemblyManifestGenerated": "2026-11-06T10:27:00.369923+00:00",
    "assemblyManifestSHA256": "dc1a4b827a088ee36a84db7e5faa753ece51843e4a9553a7b38a1d6f7be032b9",
    "assemblyPartCount": 83,
    "allAssemblyMeshHashesMatch": true,
    "nativeExportAuditGenerated": "2026-11-06T10:25:48.995958+00:00",
    "nativeExportAuditSHA256": "9772376714f84fc84d1c7f96296410a17316480f3f1ae5e37dd2101fa2f58b4b",
    "currentWrappersPass": true,
    "note": "The actual Opus report describes supervisor wrappers as stale at its initial read. They have since been refreshed by the mechanical handoff coordinator and independently hash-checked here. Vendor detailed reference shapes are deliberately excluded from the public assembly; fabricated geometry is unchanged. No physical validation performed."
  }
}
