Ty_Tech/SlabUploader
6
0
Fork 0
Code Issues 15 Pull Requests Actions Packages Projects Releases Wiki Activity
Files
6e918db3a554ef00062c7b7db890022185eb4862
SlabUploader/docs/PM-DESIGN-ALIGNMENT.md
T

24 KiB
Raw Blame History

SlabUploader — PM requirements ↔ design alignment report

Status: WORKSPACE ONLY · 2026-08-31 PT · do not commit/push (Ty HOLD on gitea writes)
Authors: Principal PM (product alignment §§1–9 + matrix) · Dev Manager (technical consistency / landmines)
Baseline commits on gitea-atd main: 457711b (AGENTS + Phase 0), 4ec478f (PRD/README/Woo/DEPLOYMENT patches)
Also compared: /workspace/slabuploader-poc-review-package.md §§1–2 (env-agnostic §3 draft not on gitea)

How to read

Label Meaning
Agreed Review package intent, AGENTS, and design specs (TECH-SPEC / ARCHITECTURE / landed patches) say the same thing
Conflict Two or more live SoT docs disagree; coding agents can pick the wrong one
Missing Required clarity absent, or only lives in /workspace drafts not on gitea

1. POC outcome & Woo firebreak
Agreed Phone → customer-ready listing on www.whidbeywoodstore.com. UAT on PROD as Woo draft + SLAB-UAT-* (forced draft). Final POC = one real SKU → Woo publish after Ty review. Settings field woo_create_status: `draft
Conflict None material on the firebreak itself after 4ec478f.
Missing Env-agnostic deploy wording still only in /workspace package §3 — not on gitea (HOLD). Does not block Woo alignment.

2. Hybrid vs server pipeline
Agreed ARCHITECTURE + TECH-SPEC + AGENTS §1 + IMPL-PLAN Phase 0 (457711b) + README + PRD stack/calibration rewrite (4ec478f): client owns mask/sliders/axis/sqft/bdft/3:4 PNG; server stores, U2Net on demand, Woo, inference proxy. Server must not recompute happy-path math as SoT.
Conflict IMPL-PLAN Phases 1+ still carry stale offline/OCR/ruler/pending/processing→ready language (banners say ignore). OPENAPI PUT still says server re-runs derived math (see Dev Manager B1).
Missing Full rewrite of IMPL-PLAN Phases 1–6 to hybrid + Woo draft UAT.

3. bdft formula
Agreed TECH-SPEC: bdft = sqft × thickness_in (TV-3). AGENTS, IMPL-PLAN Phase 0, PRD FR4 (patched), DEPLOYMENT smoke.
Conflict None on formula after 4ec478f. CONTENT-WOO length_ft = floor(length_in / 12) is display conversion, not bdft.
Missing —

4. Image prep (3:4, 1600, sliders)
Agreed TECH-SPEC §§3–5: sheet/sliders, mask confirm, 3:4 PNG, 80% fill, ≥1600 or retake, no fake upscale. AGENTS + ARCHITECTURE + review package §2 match.
Conflict Narrow: PRD FR19 still “species and wood category” pricing while TECH-SPEC/DATA-MODEL = species-only $/bdft. CONTENT-WOO media “JPEG” vs TECH-SPEC transparent PNG (Dev Manager W4).
Missing Explicit “wood category for taxonomy, not for $/bdft in POC” in PRD when writes reopen.

5. Secrets location
Agreed ARCHITECTURE + AGENTS §4 + PRD D15 (patched) + DATA-MODEL: Woo + inference secrets server AES-GCM; SLAB_AES_KEY env only; never browser.
Conflict Soft: DATA-MODEL/OPENAPI still list sheet/slider prefs in server settings while ARCHITECTURE/AGENTS say localStorage (Dev Manager W3).
Missing —

6. Inference OFF-first, then Call 1/2
Agreed Review package + AGENTS §6 + ARCHITECTURE + PROMPTS + Phase 0 out-of-scope: suite passes OFF; Gate C = OFF + manual + Woo draft; Call 1/2 after Gate C, still inside POC; Call 2 user-triggered; LoRA deferred; no LLM dimensions.
Conflict None on sequencing intent.
Missing Phase 5 rewrite later for “all photos @1024” vs older “top-down only” flavor if any remains.

7. Slab lifecycle vs woo_create_status
Agreed Slab: `draft → calibrated → ready → publishing → published
Conflict Soft word overload on “published.” OPENAPI POST “calibrated draft” returning status: draft (Dev Manager W1).
Missing Optional glossary row when writes reopen.

8. Online-only
Agreed Review package + AGENTS + README (4ec478f) + PRD decision 18 + TECH-SPEC. Phase 0 excludes offline PWA.
Conflict IMPL-PLAN Phase 2 / Phase 6 still specify offline capture (superseded banners — agents may skim past).
Missing Phase 2/6 body rewrite when writes reopen.

9. Phase / Gate sequencing
Agreed Phase 0–6 retained. Phase 0 = Gates A+B. Gate C = early Phase 1. Inference after Gate C.
Conflict Gate A raced onto gitea while Ty HOLDed further writes and asked env-agnostic + this alignment. Deploy host hard-coding still in AGENTS §9 / DEPLOYMENT / PRD FR42 (B2/B5).
Missing Ty sign via CoS on env-agnostic deploy amend before Gate B. Coding agents / inventory held.

Matrix (quick)

Topic Status
POC outcome & Woo firebreak Agreed (on gitea)
Hybrid vs server pipeline Agreed core; Conflict residual Phase 1+ / OPENAPI PUT
bdft formula Agreed
Image prep Agreed; narrow pricing/PNG conflicts
Secrets location Agreed; soft prefs storage drift
Inference OFF → Call 1/2 Agreed
Slab lifecycle vs woo_create_status Agreed (wording overload residual)
Online-only Agreed; Phase 2/6 text stale
Phase/Gate sequencing Agreed map; Missing env-agnostic sign before Gate B

Principal PM conclusion

Core product intent and design SoT (TECH-SPEC, ARCHITECTURE, AGENTS, Phase 0, Woo firebreak) align after 457711b/4ec478f. Remaining risk is implementer landmines (OPENAPI recompute, hard-coded host, stale IMPL-PLAN bodies) — not a PM↔architecture split.

Recommend via CoS: when HOLD lifts, authorize one small docs commit for env-agnostic deploy + OPENAPI “no server recompute SoT” before any coding agents or Gate B; optional IMPL-PLAN 1+/2/6 and PRD FR19 cleanup in the same or follow-on commit.

Dev Manager — technical consistency (full audit)

SlabUploader PM↔design alignment — technical consistency

Status: Workspace-only audit (HOLD). No gitea commit/push/SSH performed.
Authors: Dev Manager (executor) · for CoS / Principal PM
Repo HEAD: 4ec478f on main (Ty_Tech/SlabUploader @ gitea-atd)
Audit date: 2026-08-31 PT
HOLD note: Do not write to gitea until CoS lifts HOLD. Fixes below are recommended workspace drafts only — this report does not edit the repo. Env-agnostic deploy amend remains unsigned (see /workspace/slabuploader-poc-review-package.md §3 / /workspace/slabuploader-env-agnostic-doc-outline.md).

1. What Gate A actually locked in repo (SHAs + files)

git log -5 --oneline
4ec478f docs: finish Gate A contradiction patches
457711b docs: Gate A freeze — AGENTS.md + hybrid Phase 0
a94d81d docs: set gitea-atd Ty_Tech/SlabUploader as remotes SoT
4d9f307 docs: lock inference review — prompts, Call 1/2, feat tags
4e97770 docs: lock revised capture pipeline and hybrid architecture
SHA Commit Files locked
457711b Gate A freeze — AGENTS.md + hybrid Phase 0 AGENTS.md (new, 148 lines); docs/IMPL-PLAN.md Phase 0 rewrite (+ banner that Phases 1+ are still stale)
4ec478f Finish Gate A contradiction patches README.md, Slab capture app PRD.txt, docs/CONTENT-WOO.md, docs/DATA-MODEL.md, docs/DEPLOYMENT.md, docs/OPENAPI.md

Already consistent before Gate A (not rewritten in those SHAs, but SoT per AGENTS precedence):

  • docs/TECH-SPEC-PIPELINE.md — client-authoritative math, bdft = sqft × thickness_in, TV-1…TV-10, no /12
  • docs/ARCHITECTURE.md — hybrid topology, AES-GCM secrets, slab status machine
  • docs/PROMPTS.md — Call 1/2 (not re-audited deeply; no Gate A delta)

Gate A product locks now in-repo (via AGENTS + patches):

  • Hybrid: client = happy-path measure/crop SoT; server = store / on-demand U2Net / Woo / inference proxy
  • bdft = sqft × thickness_in (no /12)
  • Secrets: server AES-GCM only; never browser
  • Online-only POC (offline/OCR/ruler deferred)
  • woo_create_status = draft | publish, default draft; SLAB-UAT-* force Woo draft on PROD
  • Health: GET /api/health
  • Phase 0 = Gates A+B; Gate C = early Phase 1 claimable Woo draft, inference OFF

2. Consistent (safe for agents)

Agents that follow AGENTS.md §0 precedence and ignore flagged-stale text will get these right:

Topic Where it agrees
Doc precedence AGENTS §0: TECH-SPEC → ARCHITECTURE → AGENTS → DATA/OPENAPI/CONTENT/PROMPTS/DEPLOY → IMPL Phase 0 only → PRD/README intent
Hybrid who-runs-what AGENTS §1, ARCHITECTURE §§1–4, TECH-SPEC §1.5 / §9.8, README Quick orientation, PRD Architecture rewrite (post-4ec478f), DEPLOYMENT layout note (no server happy-path SoT)
bdft formula + TV-3 TECH-SPEC §4.6 + TV-3; AGENTS §3; PRD FR4 patched; IMPL Phase 0; DEPLOYMENT smoke §8.4
No fake upscale / ≥1600 TECH-SPEC §5 + TV-6/7; AGENTS §2–3; IMPL Phase 0
Secrets AES-GCM server ARCHITECTURE §5; AGENTS §4; DATA-MODEL settings; OPENAPI SettingsView masks secrets; PRD D15 patched
Woo firebreak AGENTS §5; CONTENT-WOO §2.4; DATA-MODEL woo_create_status; OPENAPI SettingsView; DEPLOYMENT smoke §8; README; PRD D12/FR36
Slab vs Woo status AGENTS §7; ARCHITECTURE §6; DATA-MODEL status machine; CONTENT-WOO §2.4 explicit “do not conflate”
Inference sequenced AGENTS §6; ARCHITECTURE clarifications 4–5; OPENAPI infer endpoints; Gate C = OFF
Phase 0 hybrid scope IMPL-PLAN Phase 0 (post-457711b) matches AGENTS; Phase 2 banner says SUPERSEDED
Health path AGENTS §9; DEPLOYMENT §7–8; OPENAPI Health; IMPL Phase 0

Bottom line for scaffold agents: If they start at AGENTS + TECH-SPEC + ARCHITECTURE + IMPL Phase 0 only, Gate B skeleton is coherent. The landmines are in Phases 1+/risks/PRD leftovers/OPENAPI one-liners and the hard-coded .201 deploy story.

3. Remaining contradictions / landmines

Severity: blocker = will cause wrong code if followed; warn = likely wrong path / wasted work; nit = wording drift, low blast radius.

BLOCKER

B1. OPENAPI: “Server re-runs derived math when inputs change”

  • Where: docs/OPENAPI.md → PUT /api/v1/slabs/{id} bullet (~L154)
  • What agents might do wrong: Implement server-side recompute of sqft/bdft/widths as SoT on PUT — directly violates AGENTS §3 / TECH-SPEC §1.5 / §9.8 (“server stores the numbers the client sends”).
  • Recommended fix (draft only): Change to: server stores client-sent derived fields; may validate shape/ranges; must not recompute mask/sqft/bdft as authoritative. Client re-runs TECH-SPEC math and PUTs new numbers.

B2. IMPL-PLAN Phase 1 invents processing→ready status

  • Where: docs/IMPL-PLAN.md Phase 1 (~L75): “Status state machine + in-process job queue (processing→ready).”
  • What agents might do wrong: Add a processing slab status and a server job that “processes” photos toward ready — reintroduces server happy-path pipeline under a new name.
  • Recommended fix: Rewrite Phase 1 to hybrid: CRUD + upload stores client-calibrated payloads; status machine = draft → calibrated → ready → publishing → published|failed only; queue only for publish/inference/U2Net-on-demand — never for measure/crop.

B3. IMPL-PLAN Phase 2 body still specifies offline/OCR/ruler (banner insufficient for careless agents)

  • Where: docs/IMPL-PLAN.md Phase 2 (~L86–106). Banner says SUPERSEDED, but Goal/Deliverables/Acceptance/Exit still demand Tesseract OCR, OpenCV.js ruler, Workbox, IndexedDB, airplane-mode UAT.
  • What agents might do wrong: Implement the body (agents often skim past blockquotes). Builds the exact “do not build” list in AGENTS §2.
  • Recommended fix: Replace Phase 2 body with a short stub: “Deferred post-POC. Capture UX lives in early Phase 1 / Gate C (online hybrid). See AGENTS §2.” Delete OCR/ruler/offline acceptance lines.

B4. IMPL-PLAN Phase 6 UAT still requires offline capture + OCR + “server normalization”

  • Where: docs/IMPL-PLAN.md Phase 6 (~L181–183, L175–179)
  • What agents might do wrong: Write UAT scripts / perf budgets for OCR 1–2s and server normalization 3–5s; gate exit on offline capture — contradicts online-only POC and client SoT.
  • Recommended fix: UAT suite = hybrid online path; drop offline capture requirement; perf = client measure/crop budgets + U2Net-on-demand only; inventory host via APP_* (see §4).

B5. Deploy SoT still hard-codes .201 / slab.tyubumini.local (Gate B blocker per review package)

  • Where: AGENTS.md §9; docs/DEPLOYMENT.md title/target/§1/§4/§6/§7; README.md L6; docs/IMPL-PLAN.md Phase 0 Gate B (.201 inventory), Phase 6, Risks, DoD; PRD FR42 / deploy bullets
  • What agents might do wrong: Bake 192.168.1.201 / slab.tyubumini.local into compose, Caddy, health URLs, runbooks; treat TyUBUMini as mandatory; skip inventory worksheet; block on wrong host.
  • Recommended fix: Apply workspace outline /workspace/slabuploader-env-agnostic-doc-outline.md — APP_HOST / APP_HOSTNAME / ports; deploy/INVENTORY.md; historical lab as non-normative callout only. Do not push until CoS lifts HOLD.

WARN

W1. OPENAPI create returns slab status draft for a “calibrated draft”

  • Where: docs/OPENAPI.md POST /api/v1/slabs (~L146–147): “Create a calibrated draft” → Returns 201 with status: draft
  • What agents might do wrong: Skip calibrated entirely; conflate “photos taken” draft with “mask+numbers uploaded” calibrated; break Gate C poll/UI assumptions.
  • Recommended fix: POST after client confirm → 201 with status: calibrated (align ARCHITECTURE §6 / DATA-MODEL / AGENTS §7). Keep local pre-upload client state out of API or document separately.

W2. AGENTS says “TV-10 only after TECH-SPEC defines it” — but TECH-SPEC already has TV-10

  • Where: AGENTS.md §3 (~L80) vs docs/TECH-SPEC-PIPELINE.md §8 TV-10 Rounding (~L192–196)
  • What agents might do wrong: Skip rounding unit tests (length 1/8", sqft/bdft 2dp, price half-up) thinking TV-10 is undefined.
  • Recommended fix: AGENTS: “Port TV-1…TV-10.” IMPL Phase 0 acceptance: include TV-10.

W3. DATA-MODEL / OPENAPI put sheet prefs in server settings; ARCHITECTURE/AGENTS say localStorage

  • Where: DATA-MODEL settings keys user_sensitivity|edge_offset|feather|sheet_mode (~L184); OPENAPI SettingsView (~L132–133); vs AGENTS §4 / ARCHITECTURE §5 (browser localStorage only)
  • What agents might do wrong: Dual-write prefs to SQLite AES settings table; or encrypt non-secrets; or fight over which wins on load.
  • Recommended fix: Remove sheet/slider prefs from server settings schema (or mark optional sync later). Keep localStorage as SoT for POC.

W4. CONTENT-WOO media upload says “JPEG bytes”; TECH-SPEC output is transparent PNG

  • Where: docs/CONTENT-WOO.md §2.4 (~L114): POST /media with JPEG bytes; TECH-SPEC §5 / PRD FR31: transparent PNG
  • What agents might do wrong: Re-encode processed PNGs to JPEG (lose alpha) before Woo media upload.
  • Recommended fix: “PNG bytes (processed inventory images)”; mention Woo accepts image/*; keep alpha.

W5. CONTENT-WOO tag auto-assign still says [character] / optional [species] tags

  • Where: docs/CONTENT-WOO.md §2.3 (~L81–83) vs DATA-MODEL / PRD D8: fig-* + feat-* only; Call 1 maps observations → feat-*
  • What agents might do wrong: Create/assign non-fig/feat Woo tags named after character strings; ignore feat-* mapping.
  • Recommended fix: Rewrite §2.3 to match DATA-MODEL: fig-* (1+) + feat-* (0+ from Call 1 mapping); no generic character/species tag invent.

W6. PRD FR17 vs D8 / DATA-MODEL on tags

  • Where: PRD FR17 “Tags: fig- only*, 1+ required” vs Decisions log #8 + DATA-MODEL fig_tag_ids + feat_tag_ids
  • What agents might do wrong: Drop feat-* from schema/UI/publish payload.
  • Recommended fix: FR17 → fig-* (1+) and feat-* (0+); cite DATA-MODEL.

W7. PRD FR19 + pricing schema still “species and wood category”

  • Where: PRD FR19 (~L134); PRD pricing schema wood_category: string (~L241); vs TECH-SPEC §6 / DATA-MODEL pricing_rules = species only flat $/bdft; AGENTS kills figure-/category-weighted pricing for v1
  • What agents might do wrong: Build wood-category pricing tiers or composite keys.
  • Recommended fix: FR19 + schema → species-only; wood-category pricing = deferred/v1.5.

W8. PRD FR40 + NFR still “server-side processing” / “server normalization”

  • Where: PRD FR40 (~L190); NFR Performance (~L200–201)
  • What agents might do wrong: Put measure/crop on publish path server-side; budget 3–5s server normalization as a feature.
  • Recommended fix: FR40 = Woo sync + status/log only (client already did processing). NFR = client local processing + optional U2Net latency; drop “server normalization” phrase.

W9. DEPLOYMENT §9 / Risks still RemBG, ruler, output_px drop to 1000

  • Where: docs/DEPLOYMENT.md §9 (~L161) RemBG/GPU on .201; docs/IMPL-PLAN.md Risks (~L201–202) ruler + output_px can drop to 1000
  • What agents might do wrong: Add RemBG server path; lower OUTPUT_PX_MIN below 1600 (violates TECH-SPEC TV-7 / no inventing pixels — undersized → retake, not lower min).
  • Recommended fix: Risks: delete ruler row for POC or mark deferred; U2Net-on-demand only (not RemBG as SoT); never drop below 1600 — retake.

W10. Caddy fragment shape is foot-gunny

  • Where: docs/DEPLOYMENT.md §4 — site-level reverse_proxy frontend:80 plus handle /api/* { reverse_proxy fastapi:8000 }
  • What agents might do wrong: Ship as-written; /api/* may never reach fastapi depending on Caddy directive order; health smoke fails.
  • Recommended fix: Use explicit handle /api/* → fastapi and handle { reverse_proxy frontend } (or handle_path) so API wins; keep service name fastapi consistent with compose.

W11. Calibration-photo language leftovers in CONTENT-WOO

  • Where: CONTENT-WOO §2.4 (~L116–117) “Calibration photos are excluded (FR2)” — PRD D1 says no calibration photos exist
  • What agents might do wrong: Invent kind=calibration in upload meta / schema (OPENAPI/DATA-MODEL correctly only inventory).
  • Recommended fix: Delete sentence; say “only inventory photos (kind=inventory).”

NIT

N1. DATA-MODEL / OPENAPI still say “PWA”

  • Where: DATA-MODEL L7 “so the PWA can reference…”; OPENAPI Versioning L301 “The PWA pins…”
  • Risk: Agents scaffold Workbox “because the contract says PWA.”
  • Fix: “client” / “SvelteKit app”; offline PWA deferred.

N2. PRD section header still “Client (mobile web / PWA)”

  • Where: PRD Architecture (~L210) — body correctly says offline deferred, but header says PWA.
  • Fix: “Client (mobile web / SvelteKit)”.

N3. ARCHITECTURE status line “Remaining sections TBD”

  • Where: ARCHITECTURE L4 — topology is locked; stale “TBD” undercuts confidence.
  • Fix: “Hybrid topology locked Gate A; remaining ops sections TBD.”

N4. TECH-SPEC status date still 2026-08-27; AGENTS Gate A 2026-08-31

  • Nit only — content matches. Optional status bump.

N5. Review package Appendix C checklist items 1–7 largely done in-repo; item 0 (env-agnostic) not done

  • Package preamble still reads like pre-Gate-A in places (“AGENTS referenced, file missing”) — workspace draft drift vs 457711b/4ec478f. Mark appendix C items 1–7 done; keep item 0 open.

N6. DEPLOYMENT Gate A banner line-break glitch

  • Where: docs/DEPLOYMENT.md L4–8 — “LAN/Tailscale” sentence split across Gate A callout (“only — Decision”). Readable but ugly; fix when env-agnostic rewrite lands.

4. Env-agnostic gap (workspace drafts vs repo still saying .201)

Artifact Role Normative host?
Repo AGENTS.md §9, docs/DEPLOYMENT.md, README.md, IMPL-PLAN.md Gate B/Phase 6/DoD, PRD FR42 Current SoT agents will read Yes — hard-codes .201 / TyUBUMini / slab.tyubumini.local / 192.168.1.201
Workspace-only /workspace/slabuploader-poc-review-package.md §3 Proposed SoT for deploy; not in repo No — APP_HOST / APP_HOSTNAME / inventory worksheet; lab example non-normative
Workspace-only /workspace/slabuploader-env-agnostic-doc-outline.md Patch outline for CoS-signed amend Same

Explicit: these two markdown files are workspace drafts, not SoT yet. Agents must not treat /workspace/*.md as repo authority. Until the env-agnostic amend lands under CoS HOLD lift, Gate B inventory step is blocked if agents assume .201 is mandatory — and equally blocked if they invent a different host without inventory.

Product locks (hybrid, bdft, Woo firebreak) are independent of this gap and already in-repo.

5. Recommended next doc patches when CoS lifts HOLD (ordered, small)

Docs-only. No scaffold/code. No reopen of hybrid/Woo/bdft locks.

  1. Env-agnostic deploy cut (Gate B unblocker) — per /workspace/slabuploader-env-agnostic-doc-outline.md:
    • AGENTS.md §9 → inventory-first + https://${APP_HOSTNAME}/api/health; lab example callout only
    • docs/DEPLOYMENT.md title/target/§1→deploy/INVENTORY.md worksheet; Caddy ${APP_HOSTNAME}; fix handle order (W10); drop RemBG SoT smell (W9)
    • README.md deploy line; PRD FR42 / deploy bullets; IMPL-PLAN Gate B / Phase 6 / Risks / DoD host wording
  2. OPENAPI one-liners (B1, W1) — PUT does not recompute SoT; POST returns calibrated
  3. AGENTS TV-10 (W2) — require TV-1…TV-10
  4. IMPL-PLAN Phase 1 (B2) — kill processing→ready; hybrid CRUD/store language
  5. IMPL-PLAN Phase 2 (B3) — replace body with deferred stub (keep banner or delete section content)
  6. IMPL-PLAN Phase 6 + Risks (B4, W9) — online UAT only; no OCR/offline/server-normalization budgets; no output_px→1000
  7. CONTENT-WOO (W4, W5, W11) — PNG media; fig-/feat- assign; no calibration photos
  8. DATA-MODEL / OPENAPI settings (W3, N1) — drop server sheet prefs (or defer); “client” not “PWA”
  9. PRD leftovers (W6–W8, N2) — FR17 feat-*; FR19 species-only; FR40/NFR hybrid wording; pricing schema
  10. Refresh workspace review package — mark Appendix C items 1–7 done; keep item 0 + sign-off for env-agnostic only (avoid agents re-doing Gate A)

After patches: single docs commit on gitea-atd; report SHA to CoS; then Gate B inventory on designated host — still no silent .201 assumption.

Audit verdict (blunt)

Gate A did land the critical product locks (hybrid, bdft, AES, Woo draft/SLAB-UAT-*, Phase 0 rewrite, AGENTS precedence). Scaffolding Gate B math/AES/health is safe if agents obey AGENTS + TECH-SPEC + Phase 0.

Gate B deploy and any agent that reads IMPL Phase 1/2/6, OPENAPI PUT, or PRD FR40/NFR without the precedence discipline will still ship wrong architecture (server recompute, processing status, offline/OCR/ruler, JPEG-without-alpha, hard-coded .201).

Highest-ROI doc fixes when HOLD lifts: (1) env-agnostic deploy, (2) OPENAPI “re-runs derived math”, (3) gut IMPL Phase 2/6 stale bodies, (4) Phase 1 status machine.

Reference in New Issue View Git Blame Copy Permalink