Ty_Tech/SlabUploader
6
0
Fork 0
Code Issues 15 Pull Requests Actions Packages Projects Releases Wiki Activity

fix(m3): AI-assist scope + confidence gate + grouped wood categories

Browse Source 
- inference: _taxonomy_lists hands the model the synced species leaves + wood
  leaves + case-insensitive attribute terms (was furniture roots / empty)
- inference: CALL1_SEED pins exact flat schema + *_confidence keys, drops grade
- inference: _gate maps field->confidence key correctly (was species_id_confidence),
  reads nested shape as safety net, omits grade (METADATA-40)
- review.svelte: figure gets AI assist, grade dropped from AI (owner judgment)
- MultiSelect: grouped mode (parents non-selectable headers, selectable leaves)
- METADATA-40 updated: grade not suggested by Call 1
This commit is contained in:
Ty
2026-09-20 21:47:36 -07:00
parent e36979617b
commit cec03c5bd3
5 changed files with 212 additions and 82 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/inference.py
+83 -23
View File
@@ -20,12 +20,18 @@ from .crypto import decrypt
router = APIRouter(prefix="/api/v1")
CALL1_SEED = (
"Classify the slab's species, wood categories, edge type, figure, and grade from the "
"supplied photos, choosing only from the supplied synced taxonomy lists, and return the "
"result fields as JSON: species_id, wood_category_ids, edge_type_term_id, figure_term_ids, "
"grade_term_id, character, inclusions, voids, checks. Report a confidence (0-1) for each "
"classified field. Map free-text observations to the supplied feature-tag list only. "
"Invent no category or term."
"Classify the slab's species, wood categories, edge type, and figure from the supplied "
"photos, choosing ONLY from the supplied synced taxonomy lists (invent no category, term, "
"or id outside them). Map free-text observations (character, inclusions, voids, checks) to "
"the supplied feature-tag list only. Grade is NOT suggested — leave it out.\n"
"Return a single JSON object with EXACTLY these top-level keys, in this shape:\n"
"{\"species_id\": <id or null>, \"species_confidence\": <0-1>, "
"\"wood_category_ids\": [<id>, ...], \"wood_category_confidence\": <0-1>, "
"\"edge_type_term_id\": <id or null>, \"edge_confidence\": <0-1>, "
"\"figure_term_ids\": [<id>, ...], \"figure_confidence\": <0-1>, "
"\"character\": [...], \"inclusions\": [...], \"voids\": [...], \"checks\": [...]}\n"
"Each *_confidence is a top-level sibling of its field (do NOT nest confidences under a "
"'confidence' object). Use 0-1 numbers; set a field to null / [] when you cannot identify it."
)
CALL2_SEED = (
"Write the long product description for the slab from the supplied photos and listing "
@@ -228,14 +234,49 @@ async def call1(slab_id: str) -> dict:
def _taxonomy_lists(conn) -> dict:
out = {"species_leaves": [], "edge_terms": [], "figure_terms": [], "grade_terms": [],
"thickness_terms": [], "moisture_terms": [], "feat_tags": []}
for r in conn.execute("SELECT id,name FROM woo_categories WHERE parent_id IS NULL").fetchall():
out["species_leaves"].append({"id": r[0], "name": r[1]})
for attr, key in (("Edge Type", "edge_terms"), ("Figure", "figure_terms"), ("Grade", "grade_terms"),
("Thickness", "thickness_terms"), ("Moisture", "moisture_terms")):
# METADATA-10: categories are leaves only; parents group only. Resolve groupers BY SLUG
# (never hardcoded id, QUALITY-105) and match attribute names case-insensitively (the
# live Woo attribute is 'Edge type', not 'Edge Type').
out = {"species_leaves": [], "wood_category_leaves": [], "edge_terms": [], "figure_terms": [],
"grade_terms": [], "thickness_terms": [], "moisture_terms": [], "feat_tags": []}
# species = leaves under the 'species' grouper
sp = conn.execute("SELECT id FROM woo_categories WHERE slug='species'").fetchone()
if sp:
for r in conn.execute(
"SELECT id,name FROM woo_categories WHERE parent_id=?", (sp[0],)
).fetchall():
out["species_leaves"].append({"id": r[0], "name": r[1]})
# wood categories = leaves under the 'wcat' grouper (leaves may sit under sub-groupers).
wc = conn.execute("SELECT id FROM woo_categories WHERE slug='wcat'").fetchone()
if wc:
kids = [r[0] for r in conn.execute(
"SELECT id FROM woo_categories WHERE parent_id=?", (wc[0],)
).fetchall()]
leaf_ids = []
for cid in kids:
has_kids = conn.execute(
"SELECT 1 FROM woo_categories WHERE parent_id=? LIMIT 1", (cid,)
).fetchone()
if has_kids:
leaf_ids.extend(r[0] for r in conn.execute(
"SELECT id FROM woo_categories WHERE parent_id=?", (cid,)
).fetchall())
else:
leaf_ids.append(cid)
if leaf_ids:
q = ",".join("?" * len(leaf_ids))
for r in conn.execute(
"SELECT id,name FROM woo_categories WHERE id IN (%s)" % q, leaf_ids
).fetchall():
out["wood_category_leaves"].append({"id": r[0], "name": r[1]})
# attribute terms, case-insensitive on the attribute name. Grade stays in the synced set
# (owner reference) but the model is told not to suggest it (METADATA-40).
for attr, key in (("Edge type", "edge_terms"), ("Figure", "figure_terms"),
("Grade", "grade_terms"), ("Thickness", "thickness_terms"),
("Moisture", "moisture_terms")):
rows = conn.execute(
"SELECT t.id,t.name FROM woo_attribute_terms t JOIN woo_attributes a ON t.attribute_id=a.id WHERE a.name=?",
"SELECT t.id,t.name FROM woo_attribute_terms t JOIN woo_attributes a ON t.attribute_id=a.id "
"WHERE lower(a.name)=lower(?)",
(attr,),
).fetchall()
out[key] = [{"id": r[0], "name": r[1]} for r in rows]
@@ -257,24 +298,43 @@ def _parse_json(raw: str) -> dict:
def _gate(parsed: dict) -> dict:
"""Apply the 0.7 confidence gate; below-threshold fields are omitted (METADATA-45)."""
def keep(field: str, conf_key: str) -> Any:
conf = parsed.get(conf_key)
"""Apply the 0.7 confidence gate; below-threshold fields are omitted (METADATA-45).
Confidence is read from the flat '<field>_confidence' keys; as a safety net it also
understands the nested 'confidence: {<field>: n}' shape some models emit, so a shape
wobble cannot silently nuke every suggestion. Grade is not suggested (METADATA-40).
"""
# Map each field to its confidence key. The field name and the confidence key do NOT
# share a stem (e.g. field 'species_id' -> key 'species_confidence'), so derive it
# explicitly rather than f"{field}_confidence".
conf_key = {
"species_id": "species_confidence",
"wood_category_ids": "wood_category_confidence",
"edge_type_term_id": "edge_confidence",
"figure_term_ids": "figure_confidence",
}
conf_obj = parsed.get("confidence") if isinstance(parsed.get("confidence"), dict) else {}
def _conf(field: str) -> Any:
k = conf_key.get(field, f"{field}_confidence")
if k in conf_obj:
return conf_obj[k]
return parsed.get(k)
def keep(field: str) -> Any:
conf = _conf(field)
if isinstance(conf, (int, float)) and conf >= _THRESHOLD and field in parsed:
return parsed[field]
return None
out: dict[str, Any] = {}
if v := keep("species_id", "species_confidence"):
if v := keep("species_id"):
out["species_id"] = v
if v := keep("wood_category_ids", "wood_category_confidence"):
if v := keep("wood_category_ids"):
out["wood_category_ids"] = v
if v := keep("edge_type_term_id", "edge_confidence"):
if v := keep("edge_type_term_id"):
out["edge_type_term_id"] = v
if v := keep("figure_term_ids", "figure_confidence"):
if v := keep("figure_term_ids"):
out["figure_term_ids"] = v
if v := keep("grade_term_id", "grade_confidence"):
out["grade_term_id"] = v
# grade intentionally not gated/suggested (METADATA-40: owner judgment).
for k in ("character", "inclusions", "voids", "checks"):
if parsed.get(k):
out[k] = parsed[k]
@@ -361,4 +421,4 @@ async def test_inference() -> dict:
except InferenceError as e:
return {"ok": False, "mode": "fail", "vision_capable": False, "detail": str(e)}
detail = f"reachable in {int((time.time()-t0)*1000)}ms"
return {"ok": True, "mode": "stateless", "vision_capable": True, "detail": detail}
return {"ok": True, "mode": "stateless", "vision_capable": True, "detail": detail}
backend/tests/test_slabs_api.py
+22
View File
@@ -201,6 +201,28 @@ def test_create_null_sku_allowed(client, db):
assert resp.json()["sku"] is None
def test_gate_reads_flat_confidence_keys(client, db):
# Regression: the gate's confidence keys are 'species_confidence' / 'edge_confidence'
# etc., NOT f'{field}_confidence' (field 'species_id' would give the wrong key). A
# high-confidence response must surface its picks; low-confidence ones must not.
from backend.inference import _gate, _parse_json
good = '```json\n{"species_id": 284, "species_confidence": 0.85, "wood_category_ids": [251], "wood_category_confidence": 1.0, "edge_type_term_id": 323, "edge_confidence": 0.95, "figure_term_ids": [332], "figure_confidence": 0.9, "character": [], "inclusions": ["bark"], "voids": [], "checks": []}\n```'
out = _gate(_parse_json(good))
assert out["species_id"] == 284
assert out["wood_category_ids"] == [251]
assert out["edge_type_term_id"] == 323
assert out["figure_term_ids"] == [332]
# grade must never be gated in (METADATA-40)
assert "grade_term_id" not in out
# low-confidence fields are dropped (METADATA-45)
low = '```json\n{"species_id": 284, "species_confidence": 0.4, "wood_category_ids": [251], "wood_category_confidence": 0.8, "edge_type_term_id": 323, "edge_confidence": 0.2, "figure_term_ids": [332], "figure_confidence": 0.9}\n```'
out2 = _gate(_parse_json(low))
assert out2["wood_category_ids"] == [251]
assert out2["figure_term_ids"] == [332]
assert "species_id" not in out2
assert "edge_type_term_id" not in out2
def test_ready_gate_rejects_malformed_sku(client, db):
# REVIEW-45 enforced in the middle tier: a present-but-out-of-range/char-set SKU is not ready.
from backend.review import compute_ready, sku_valid