feat(style-acq T4+T5): פנקס-התאמה draft↔final + דיסטילציה אוטומטית דרך ה-curator
סוגר את לולאת-הלמידה (INV-LRN4): כל החלטה נסגרת מול הסופי, וכל סופי מנותח מול הטיוטה. מזין את הטבלאות ש-T15 כבר קורא מהן. T5 — פנקס-התאמה: - SCHEMA_V26: טבלת draft_final_pairs (snapshot draft + final + diff + analysis + status). - db: create/update/list_draft_final_pairs. - mark-final (app.py): תופס snapshot של הטיוטה (decision_blocks) ברגע החתימה, לפני שאפשר לדרוס אותו, ופותח שורת-פנקס (status=final_received). T4 — דיסטילציה אוטומטית: - learning_loop.process_final_version: משתמש ב-snapshot (לא בבלוקים שאולי השתנו), מסווג style_method↔substance, שומר הצעה ב-pair (status=analyzed). **הוסר ה-auto-upsert של style_patterns** — ביטל את ה-bug שדרס את שער-היו"ר וזיהם סגנון במהות (INV-LRN1 + INV-LRN5). - LESSONS_PROMPT: הפרדת style_method↔substance מפורשת + לקח מופשט בלבד. - curator wake + hermes-curator.md: מריץ ingest_final_version ראשון; מציע רק style_method שלא תועד; substance→מסלול precedent. INV-LRN1 (שער-יו"ר, אין auto-commit) · INV-LRN4 (ניגוד-אמת) · INV-LRN5 (טוהר). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
@@ -1181,6 +1181,29 @@ ALTER TABLE precedent_chunks
|
||||
ADD COLUMN IF NOT EXISTS halacha_extracted_at TIMESTAMPTZ;
|
||||
"""
|
||||
|
||||
SCHEMA_V26_SQL = """
|
||||
-- draft_final_pairs (T5 / INV-LRN4): the reconciliation ledger.
|
||||
-- Every decision is "closed" only after it is compared against the chair's signed
|
||||
-- final. Captures an immutable snapshot of the AI draft at mark-final time (before
|
||||
-- it can be overwritten), paired with the final. The LLM distillation (curator)
|
||||
-- fills final_text + diff_stats + analysis later and advances status.
|
||||
CREATE TABLE IF NOT EXISTS draft_final_pairs (
|
||||
id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
|
||||
case_id UUID NOT NULL REFERENCES cases(id) ON DELETE CASCADE,
|
||||
draft_text TEXT NOT NULL DEFAULT '',
|
||||
final_path TEXT DEFAULT '',
|
||||
final_text TEXT DEFAULT '',
|
||||
diff_stats JSONB DEFAULT NULL,
|
||||
analysis JSONB DEFAULT NULL,
|
||||
-- final_received → analyzed → lessons_folded
|
||||
status TEXT NOT NULL DEFAULT 'final_received',
|
||||
created_at TIMESTAMPTZ DEFAULT now(),
|
||||
updated_at TIMESTAMPTZ DEFAULT now()
|
||||
);
|
||||
CREATE INDEX IF NOT EXISTS idx_draft_final_pairs_case ON draft_final_pairs(case_id);
|
||||
CREATE INDEX IF NOT EXISTS idx_draft_final_pairs_status ON draft_final_pairs(status);
|
||||
"""
|
||||
|
||||
|
||||
async def _run_schema_migrations(pool: asyncpg.Pool) -> None:
|
||||
async with pool.acquire() as conn:
|
||||
@@ -1210,7 +1233,8 @@ async def _run_schema_migrations(pool: asyncpg.Pool) -> None:
|
||||
await conn.execute(SCHEMA_V23_SQL)
|
||||
await conn.execute(SCHEMA_V24_SQL)
|
||||
await conn.execute(SCHEMA_V25_SQL)
|
||||
logger.info("Database schema initialized (v1-v25)")
|
||||
await conn.execute(SCHEMA_V26_SQL)
|
||||
logger.info("Database schema initialized (v1-v26)")
|
||||
|
||||
|
||||
async def init_schema() -> None:
|
||||
@@ -2241,6 +2265,70 @@ async def get_recent_decision_lessons(limit: int = 15, practice_area: str = "")
|
||||
return [dict(r) for r in rows]
|
||||
|
||||
|
||||
async def create_draft_final_pair(case_id: UUID, draft_text: str, final_path: str = "") -> str:
|
||||
"""Capture the draft↔final pairing at mark-final (T5 / INV-LRN4). Immutable draft
|
||||
snapshot; final_text/diff_stats/analysis filled later by the curator distillation."""
|
||||
pool = await get_pool()
|
||||
async with pool.acquire() as conn:
|
||||
row = await conn.fetchrow(
|
||||
"""INSERT INTO draft_final_pairs (case_id, draft_text, final_path, status)
|
||||
VALUES ($1, $2, $3, 'final_received') RETURNING id""",
|
||||
case_id, draft_text, final_path,
|
||||
)
|
||||
return str(row["id"])
|
||||
|
||||
|
||||
async def update_draft_final_pair(
|
||||
pair_id: UUID,
|
||||
final_text: str | None = None,
|
||||
diff_stats: dict | None = None,
|
||||
analysis: dict | None = None,
|
||||
status: str | None = None,
|
||||
) -> None:
|
||||
"""Advance a pairing row (curator distillation): final_text → diff_stats → analysis → status."""
|
||||
sets, params, idx = [], [], 1
|
||||
if final_text is not None:
|
||||
sets.append(f"final_text = ${idx}"); params.append(final_text); idx += 1
|
||||
if diff_stats is not None:
|
||||
sets.append(f"diff_stats = ${idx}::jsonb"); params.append(json.dumps(diff_stats, ensure_ascii=False)); idx += 1
|
||||
if analysis is not None:
|
||||
sets.append(f"analysis = ${idx}::jsonb"); params.append(json.dumps(analysis, ensure_ascii=False)); idx += 1
|
||||
if status is not None:
|
||||
sets.append(f"status = ${idx}"); params.append(status); idx += 1
|
||||
if not sets:
|
||||
return
|
||||
sets.append("updated_at = now()")
|
||||
params.append(pair_id)
|
||||
pool = await get_pool()
|
||||
async with pool.acquire() as conn:
|
||||
await conn.execute(
|
||||
f"UPDATE draft_final_pairs SET {', '.join(sets)} WHERE id = ${idx}", *params,
|
||||
)
|
||||
|
||||
|
||||
async def list_draft_final_pairs(status: str | None = None, limit: int = 200) -> list[dict]:
|
||||
"""Reconciliation ledger: all decisions paired with their final + status."""
|
||||
pool = await get_pool()
|
||||
async with pool.acquire() as conn:
|
||||
if status:
|
||||
rows = await conn.fetch(
|
||||
"""SELECT p.id, p.case_id, c.case_number, c.title, p.status,
|
||||
p.diff_stats, p.created_at, p.updated_at
|
||||
FROM draft_final_pairs p LEFT JOIN cases c ON c.id = p.case_id
|
||||
WHERE p.status = $1 ORDER BY p.created_at DESC LIMIT $2""",
|
||||
status, limit,
|
||||
)
|
||||
else:
|
||||
rows = await conn.fetch(
|
||||
"""SELECT p.id, p.case_id, c.case_number, c.title, p.status,
|
||||
p.diff_stats, p.created_at, p.updated_at
|
||||
FROM draft_final_pairs p LEFT JOIN cases c ON c.id = p.case_id
|
||||
ORDER BY p.created_at DESC LIMIT $1""",
|
||||
limit,
|
||||
)
|
||||
return [dict(r) for r in rows]
|
||||
|
||||
|
||||
async def upsert_style_pattern(
|
||||
pattern_type: str,
|
||||
pattern_text: str,
|
||||
|
||||
@@ -51,26 +51,25 @@ def compute_diff_stats(draft_text: str, final_text: str) -> dict:
|
||||
}
|
||||
|
||||
|
||||
LESSONS_PROMPT = """אתה מנתח שינויים בהחלטות משפטיות. קיבלת טיוטה (שנוצרה ע"י AI) וגרסה סופית (שעברה עריכת דפנה).
|
||||
LESSONS_PROMPT = """אתה מנתח את הפער בין טיוטה (AI) לגרסה סופית שדפנה תמיר חתמה, כדי ללמוד **איך דפנה כותבת ומנתחת** — לא את ההלכה הספציפית.
|
||||
|
||||
## הבחנה קריטית (INV-LRN5 — טוהר-הקול):
|
||||
לכל שינוי קבע `domain`:
|
||||
- **style_method** — *איך* דפנה כותבת/חושבת: ניסוח, קצב, מבנה, תנועות-הנמקה, ביטויי-מעבר, טון, סדר-טיפול. **זה מה שלומדים** (ניתן להכללה לכל תיק).
|
||||
- **substance** — תוכן ספציפי-לתיק: הלכה, עובדה, תקדים, מספר. **לא לומדים** (לא ניתן לגרור לתיק אחר).
|
||||
|
||||
## משימה:
|
||||
1. זהה את השינויים המהותיים (לא הקלדה/פורמט)
|
||||
2. סווג כל שינוי:
|
||||
- expression_change — ביטוי שהוחלף (הצע כלקח לעתיד)
|
||||
- structure_change — שינוי מבני (סדר, חלוקה)
|
||||
- content_addition — תוכן שנוסף (מה חסר?)
|
||||
- content_removal — תוכן שהוסר (מה מיותר?)
|
||||
- tone_change — שינוי טון (רשמי יותר/פחות)
|
||||
- error_fix — תיקון שגיאה עובדתית/משפטית
|
||||
3. הסק לקחים שניתן להפעיל בהחלטות עתידיות
|
||||
1. זהה שינויים מהותיים (לא הקלדה/פורמט/מספור-אוטומטי).
|
||||
2. לכל שינוי: `type` (expression_change / structure_change / content_addition / content_removal / tone_change / reasoning_move / error_fix) + `domain` (style_method / substance).
|
||||
3. הסק לקח **מופשט** (על השיטה/הקול, לא על התוכן) — רק עבור style_method.
|
||||
|
||||
## פלט JSON:
|
||||
{
|
||||
"changes": [
|
||||
{"type": "...", "description": "תיאור השינוי", "draft_text": "...", "final_text": "...", "lesson": "לקח לעתיד"}
|
||||
{"type": "...", "domain": "style_method|substance", "block": "block-yod", "description": "...", "draft_text": "...", "final_text": "...", "lesson": "לקח מופשט (style_method בלבד)"}
|
||||
],
|
||||
"new_expressions": ["ביטוי חדש שדפנה הוסיפה"],
|
||||
"overall_assessment": "הערכה כללית (1-2 משפטים)"
|
||||
"new_expressions": ["ביטוי-מעבר/נוסחה חדשים (style_method בלבד — לא הלכות)"],
|
||||
"overall_assessment": "1-2 משפטים"
|
||||
}
|
||||
"""
|
||||
|
||||
@@ -114,42 +113,53 @@ async def process_final_version(
|
||||
if not decision:
|
||||
raise ValueError(f"No decision for case {case_id}")
|
||||
|
||||
# Get draft text (combine all blocks)
|
||||
# Prefer the immutable snapshot captured at mark-final (T5/INV-LRN4); fall back
|
||||
# to the live blocks (which may have been edited after sign-off).
|
||||
pool = await db.get_pool()
|
||||
pair_id = None
|
||||
draft_text = ""
|
||||
async with pool.acquire() as conn:
|
||||
rows = await conn.fetch(
|
||||
"""SELECT content FROM decision_blocks
|
||||
WHERE decision_id = $1 AND word_count > 0
|
||||
ORDER BY block_index""",
|
||||
UUID(decision["id"]),
|
||||
pair = await conn.fetchrow(
|
||||
"""SELECT id, draft_text FROM draft_final_pairs
|
||||
WHERE case_id = $1 AND status = 'final_received'
|
||||
ORDER BY created_at DESC LIMIT 1""",
|
||||
case_id,
|
||||
)
|
||||
draft_text = "\n\n".join(r["content"] for r in rows if r["content"])
|
||||
if pair:
|
||||
pair_id = pair["id"]
|
||||
draft_text = pair["draft_text"] or ""
|
||||
if not draft_text:
|
||||
rows = await conn.fetch(
|
||||
"""SELECT content FROM decision_blocks
|
||||
WHERE decision_id = $1 AND word_count > 0
|
||||
ORDER BY block_index""",
|
||||
UUID(decision["id"]),
|
||||
)
|
||||
draft_text = "\n\n".join(r["content"] for r in rows if r["content"])
|
||||
|
||||
if not draft_text:
|
||||
raise ValueError("No draft content to compare")
|
||||
|
||||
# Compute stats
|
||||
# Compute stats (pure) + AI distillation (style/method vs substance)
|
||||
diff_stats = compute_diff_stats(draft_text, final_text)
|
||||
|
||||
# Analyze changes with AI
|
||||
analysis = await analyze_changes(draft_text, final_text)
|
||||
|
||||
# Store new expressions as style patterns
|
||||
for expr in analysis.get("new_expressions", []):
|
||||
if expr and len(expr) > 3:
|
||||
await db.upsert_style_pattern(
|
||||
pattern_type="characteristic_phrase",
|
||||
pattern_text=expr,
|
||||
context="למד מגרסה סופית",
|
||||
)
|
||||
# INV-LRN1: do NOT auto-commit learnings into writer-consumed channels.
|
||||
# The distillation is a PROPOSAL stored on the pair; the chair/curator approves
|
||||
# it (→ decision_lessons / appeal_type_rules, surfaced by T15) via the gate.
|
||||
# (Previously this auto-upserted every new_expression as a style_pattern —
|
||||
# that both bypassed the gate and contaminated style with substance. Removed.)
|
||||
if pair_id is not None:
|
||||
await db.update_draft_final_pair(
|
||||
UUID(str(pair_id)),
|
||||
final_text=final_text,
|
||||
diff_stats=diff_stats,
|
||||
analysis=analysis,
|
||||
status="analyzed",
|
||||
)
|
||||
|
||||
# Update decision status
|
||||
await db.update_decision(
|
||||
UUID(decision["id"]),
|
||||
status="final",
|
||||
)
|
||||
|
||||
# Update case status
|
||||
# Update decision + case status
|
||||
await db.update_decision(UUID(decision["id"]), status="final")
|
||||
case = await db.get_case(case_id)
|
||||
if case:
|
||||
await db.update_case(case_id, status="final")
|
||||
@@ -157,6 +167,7 @@ async def process_final_version(
|
||||
return {
|
||||
"diff_stats": diff_stats,
|
||||
"analysis": analysis,
|
||||
"pair_id": str(pair_id) if pair_id else None,
|
||||
"lessons_count": len(analysis.get("changes", [])),
|
||||
"new_expressions": len(analysis.get("new_expressions", [])),
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user