docs: align tracked files to single coherent SoT

Contradictions resolved:

- AGENTS.md: add 'Known stale PRD lines' section (FR17, FR19, FR40, NFR, FR42)
  so PRD's cross-reference resolves
- AGENTS §6: remove 'tag sets' from Call 1 body description (Call 1 only uses
  species and attribute lists)
- PROMPTS.md: shipped defaults path corrected to backend/app/prompts/ (matches
  DEPLOYMENT repo layout)
- PROMPTS.md: Call 1 notes removed feat-* and fig-* from taxonomy list (Call 1
  returns free-text observations mapped to feat-* tags by server, not fig-*)
- PROMPTS.md: Call 2 template field names aligned with CONTENT-WOO defaults
  (width_avg_in, width_min_in, width_max_in, length_in_remainder)
- UX.md §112: remove 'tag lists' from Call 1 body description
- DATA-MODEL: prompt files location corrected from /data/prompts/ to
  backend/app/prompts/ (bind-mounted read-only to /app/prompts/)
- DEPLOYMENT: backup script updated (prompts are version-controlled in repo,
  no separate backup needed)
- docker-compose.yml: prompts mount corrected to ../backend/app/prompts

docs/DATA-MODEL.md

@@ -194,10 +194,11 @@ Sheet/slider prefs (`sheet_mode`, sensitivity, edge offset, feather) are **not**
 settings keys in POC. Client localStorage is source of truth (AGENTS §4 / ARCHITECTURE §5).
 Optional server sync of those prefs is out of scope for POC.
 
-**Prompt files** — not in the settings table. Stored as text files in
-`/data/prompts/` (server volume). Read fresh on each invocation. Default
-prompts shipped with the app; owner edits directly. A reset-to-default script
-restores originals.
+**Prompt files** — not in the settings table. Shipped as text files in
+`backend/app/prompts/` (repo), bind-mounted read-only to `/app/prompts/` in the
+FastAPI container. Read fresh on each invocation. Default prompts shipped with the
+app; owner edits on the host; no restart needed. A reset-to-default script
+restores originals from the app bundle.
 
 This table is the ciphertext **key inventory** and schema only. The never-browser /
 AES-GCM / `SLAB_AES_KEY` env policy lives in AGENTS.md §4; flow lives in ARCHITECTURE §5.

docs/DEPLOYMENT.md

@@ -147,11 +147,14 @@ provide `python -m app.security.rekey NEW_KEY` (reads old key from env
 ```
 1. sqlite3 /data/slab.db ".backup /backups/slab-YYYYMMDD.db"   # WAL-safe
 2. tar czf /backups/slab-images-YYYYMMDD.tgz -C /images .
-3. cp -r /backups/prompts-YYYYMMDD.tar.gz prompts/              # prompt files
-4. retain: 7 daily + 4 weekly; prune older
-5. log result to /var/log/slab-backup.log; alert (Hermes) on failure
+3. retain: 7 daily + 4 weekly; prune older
+4. log result to /var/log/slab-backup.log; alert (Hermes) on failure
 ```
 
+Prompt files live in the git repo (`backend/app/prompts/`) and are version-controlled;
+they do not need a separate backup. The `reset-to-default` script restores shipped
+defaults from the app bundle.
+
 - Backup target: local dir first; mirror to an off-box share for safety (same pattern
   as other stacks on the host).
 - **Restore**: stop compose → restore DB file + untar images → start compose.

docs/PROMPTS.md

@@ -2,17 +2,17 @@
 
 Status: SPEC · 2026-08-27 · SlabUploader
 Prompts are server-side text files, not settings values. Read fresh on each
-invocation. Shipped defaults live in `backend/prompts/`.
+invocation. Shipped defaults live in `backend/app/prompts/`.
 
 ## Management
 
 - **Location:** `/app/prompts/` inside the FastAPI container, mounted read-only
-  from host `prompts/`.
+  from host `backend/app/prompts/`.
 - **Editing:** Edit the text files on the host. No server restart needed.
 - **Versioning:** The inference_log stores the prompt content (or hash) with each
   inference call, so you can trace which prompt version produced which result.
 - **Reset-to-default:** `python -m app.prompts.reset` restores shipped defaults
-  from the app bundle.
+  from `backend/app/prompts/`.
 
 ## Call 1 — Vision (species/character classification)
 
@@ -85,7 +85,7 @@ Return valid JSON matching this schema:
 ```
 
 ### Notes
-- The taxonomy list (species, wood categories, edge type, figure, grade, feat-*) is
+- The taxonomy list (species, wood categories, edge type, figure, grade) is
   pulled from the **synced Woo cache only**. No hardcoded mill species list. A term
   not in the cache cannot be suggested or selected.
 - Brief definitions help the model distinguish ambiguous cases
@@ -95,8 +95,8 @@ Return valid JSON matching this schema:
   is the authoritative source for feat-* tag definitions; the design does not
   duplicate them.
 - **Pre-population (when `inference_enabled`):** Call 1 results are the primary
-  source that fills the review screen for species, character, wood categories,
-  figure/edge/grade attributes, and mapped `fig-*`/`feat-*` tags.
+  source that fills the review screen for species, wood categories, edge type,
+  figure, grade, and mapped `feat-*` tags from free-text observations.
 - **Confidence gate (default threshold 0.7):**
   - Field confidence **≥ threshold** → pre-populate that field (mutable;
     user may clear or change it).
@@ -171,9 +171,9 @@ Return JSON: { "prose": "..." }
 The server assembles the final title and description from templates with
 deterministic measurements, inserting the LLM prose where appropriate.
 
-Example title: `{species} {length_ft}ft × {width_avg}in Slab — {figure}`
+Example title: `{species} {length_ft}ft {length_in_remainder}in × {width_avg_in}in Wood Slab — {figure}`
 Example description: `{llm_prose}. Each slab is naturally shaped — width varies
-along the length at {width_min}in to {width_max}in. Sold by the slab;
+along the length at {width_min_in}in to {width_max_in}in. Sold by the slab;
 dimensions as measured. {geo_sentence}`
 
 ### Notes

docs/UX.md

@@ -109,7 +109,7 @@ values always win and every value stays editable.
 ### What vision fills, and when
 
 When inference is on, Call 1 runs after the calibrated draft uploads. It sends all inventory photos
-downscaled to 1024 plus the store's synced species, attribute, and tag lists, and returns a guess
+downscaled to 1024 plus the store's synced species and attribute lists, and returns a guess
 with a confidence per field. The threshold is 0.7 by default:
 
 - A field at or above 0.7 pre-fills on review. It stays editable; the owner can change or clear it.