Chaim 970e8dc748 feat(storage): #106.4 — DB-driven blob→MinIO migration script (dry-run default) ...

הגירת בלובים מדיסק ל-MinIO, מונחית-DB ולא `mc mirror` גורף — כי ה-bucket נקבע
per-file-SEMANTIC (מסמך/טיוטה→documents, thumbnail→derived). סורק 6 עמודות-נתיב
שקיימות בפועל (documents.file_path · cases.active_draft_path · digests.source_document_path
· draft_final_pairs.final_path · document_image_embeddings/precedent_image_embeddings.
image_thumbnail_path) — לא כפי שהספ הניח (case_law.source_document_path/*_image_pages לא קיימים).

מטפל ב-3 פורמטי-נתיב legacy לא-עקביים (אומת 2026-06-11): container-abs `/data/…`,
host-abs `/home/chaim/legal-ai/data/…`, ו-relative — מנרמל ל-key יחסי-DATA_DIR (תואם
storage.normalize_key + אתרי-הכתיבה #106.3 + read-wiring העתידי #106.5). קבצים שלא
נמצאים/מחוץ-ל-DATA_DIR מדווחים, לא נבלעים.

dry-run (ברירת-מחדל): תוכנית + מניפסט CSV ל-data/audit, אפס-שינוי. --apply מעלה דרך mcli
ומאמת size אחרי כל PUT; **הדיסק לא נוגע** → re-run אידמפוטנטי וההגירה הפיכה (לרוקן דליות
+ flip חזרה ל-filesystem). נרמול עמודות-ה-DB ל-keys נקיים = צעד נפרד מאוחר (#106.5).

אומת חי (dry-run): derived 2593 (260MB) · documents 811 (638MB) · 0 outside · 28 חסרים
(רפרנסי-DB תלויים מראש). סה"כ 3404 קבצים / 899MB.

invariants: G2 (key=normalize_key, מסלול-אחסון יחיד) · INV-STG1/3 (storage layer, bucket
per-governance) · INV-G10 (dry-run/הפיך, לא נוגע בדיסק). הצעדים הבלתי-הפיכים (cutover/WORM)
נפרדים ועוצרים לאישור.
tests: dry-run חי = אימות (count+size+normalization). py_compile OK.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

2026-06-11 17:19:05 +00:00

..

.archive

feat(curator): switch Hermes Curator to DeepSeek V4-Pro via deepseek_local adapter

2026-05-10 05:58:52 +00:00

_pipeline_runtime.py

feat(pipeline): durable execution for final_halacha via LangGraph (P0, X16/INV-DUR1, #114)

2026-06-10 09:52:35 +00:00

ab_halacha_opus48.py

feat(spec): X11 citation-corroboration + INV-G10 amendment + Opus 4.8 halacha extraction

2026-05-31 18:42:13 +00:00

audit_corpus_integrity.py

feat: Stage C — RAG advanced (#33, #47, #48, #49, #50, #51)

2026-05-26 11:26:52 +00:00

audit_training_corpus.py

feat(training): Style Studio — upload, rich corpus, lessons, curator portrait, chat

2026-05-27 10:06:22 +00:00

auto-sync-cases.sh

Fix case repo sync + auto-create Gitea repos + add sync indicator

2026-04-14 15:28:16 +00:00

backfill_chunk_pages.py

fix(retrieval): rewrite chunk-page retrofit to skip OCR

2026-05-03 20:04:33 +00:00

backfill_legal_arguments.py

feat: Stage A finalizers + #35/#36/#37 — critical-gap closure

2026-05-26 08:34:40 +00:00

backfill_missing_precedents.py

feat(X13 Tier-0): decode supremedecisions API — fetch serial-format Supreme verdicts

2026-06-08 06:53:31 +00:00

backfill_multimodal_precedents.py

feat: Stage C — RAG advanced (#33, #47, #48, #49, #50, #51)

2026-05-26 11:26:52 +00:00

backfill_nevo_preamble.py

feat(nevo): backfill leaked preamble + ratio gold-set benchmark (#86)

2026-06-06 19:45:43 +00:00

backfill_style_exemplars.py

feat(style-acq T1-T3): קורפוס-דוגמאות של דפנה לכותב (style_exemplars)

2026-06-06 18:10:01 +00:00

backup-db.sh

Add full decision writing pipeline: classify, extract, brainstorm, write, QA, export

2026-04-03 10:21:47 +00:00

bidi_table.py

Improve document processing pipeline and agent workflows

2026-04-09 16:45:49 +00:00

calibrate_halacha_dedup.py

feat(halacha): application gate + lexical dedup tail + quality harnesses (#81,#82)

2026-06-06 19:55:45 +00:00

compute_ndcg.py

feat: Stage C — RAG advanced (#33, #47, #48, #49, #50, #51)

2026-05-26 11:26:52 +00:00

convert_decision_template.py

Pre-existing agent updates + analysis DOCX export

2026-04-16 18:49:10 +00:00

curator_apply_pipeline_branch.py

feat(cases): תצוגת "פסיקה שצוטטה בהחלטה" בעמוד-התיק + שחזור חיווט-הרמס

2026-06-08 11:59:21 +00:00

deploy-track-changes.sh

Add Track Changes architecture for draft revisions (CMP + CMPA)

2026-04-16 18:49:30 +00:00

drain_court_fetch.py

feat(ops): /operations — מוני-תור אחידים, "מה רץ עכשיו", וניהול-תהליכים

2026-06-08 08:57:23 +00:00

drain_digests.py

feat(ops): /operations — מוני-תור אחידים, "מה רץ עכשיו", וניהול-תהליכים

2026-06-08 08:57:23 +00:00

drain_halacha_queue.py

feat(halacha): drain לילי (23:00–05:00) + per-upload חילוץ תיק-בודד דרך ה-CEO (#120)

2026-06-11 14:02:38 +00:00

drain_metadata_queue.py

feat(ops): /operations — מוני-תור אחידים, "מה רץ עכשיו", וניהול-תהליכים

2026-06-08 08:57:23 +00:00

eval_gold_bootstrap.py

feat(eval): FU-5 — retrieval eval harness + halacha backlog visibility (#63)

2026-05-31 14:58:13 +00:00

eval_retrieval.py

feat(eval): FU-5 — retrieval eval harness + halacha backlog visibility (#63)

2026-05-31 14:58:13 +00:00

final_halacha_pipeline.py

feat(pipeline): durable execution for final_halacha via LangGraph (P0, X16/INV-DUR1, #114)

2026-06-10 09:52:35 +00:00

final_learning_pipeline.py

feat(pipeline): durable execution for final_learning via shared runtime (P1, X16/INV-DUR1, #115)

2026-06-10 09:57:01 +00:00

fix_paperclipai_skills_drift.py

chore(skills): remove paperclip-dev, scope converting-plans-to-tasks

2026-05-04 17:47:05 +00:00

fu2b_reconcile_internal_case_numbers.py

feat(fu2b): flag PROC_MISMATCH (case_number prefix vs proceeding_type) for chair

2026-05-31 08:57:42 +00:00

fu2c_reconcile_external_case_numbers.py

feat(migration): FU-2c — reconcile external case_law identifiers (GAP-08, #68)

2026-05-31 14:12:45 +00:00

goldset_ai_recommend.py

fix(halacha): split authority (derived) from rule_role — stop source-conflation (INV-DM7)

2026-06-07 18:18:41 +00:00

goldset_independent_judge.py

feat(goldset): independent second-judge for rule_role (break AI-anchoring)

2026-06-07 20:12:58 +00:00

goldset_panel_label.py

fix(halacha): #81.7 — report Gwet AC1 + consensus-vs-human (κ paradox under skew)

2026-06-11 16:13:24 +00:00

halacha_batch_reconcile.py

feat(halacha): equivalent-halacha (parallel-authority) links across precedents

2026-06-06 21:29:46 +00:00

halacha_goldset.py

feat(halacha): application gate + lexical dedup tail + quality harnesses (#81,#82)

2026-06-06 19:55:45 +00:00

halacha_panel_approve.py

feat(learning): חיווט אוטונומי לכפתורי מסלול-הסופי — סקריפט-תזמור אחד לכל שלב

2026-06-08 10:21:39 +00:00

halacha_panel_audit.py

feat(halacha): panel safety-net audit (selective-prediction monitoring)

2026-06-08 05:01:03 +00:00

halacha_panel_calibrate.py

feat(halacha): multi-judge approval panel + policy calibration (Trust-or-Escalate)

2026-06-07 21:11:30 +00:00

halacha_rule_role_backfill.py

fix(halacha): split authority (derived) from rule_role — stop source-conflation (INV-DM7)

2026-06-07 18:18:41 +00:00

ingest_bulletins.py

feat(bulletins): catalog monthly "עו"ד על נדל"ן" bulletins into the radar (X12)

2026-06-08 08:07:45 +00:00

ingest_digests_batch.py

refactor(digests): single source of truth — drop processed/ folder state (X12)

2026-06-07 20:33:18 +00:00

leak_guard.py

feat(ci): G12 leak-guard — enforce the Agent Platform Port seam (R4, #113)

2026-06-10 09:40:42 +00:00

legal-chat-service.config.cjs

docs: record Infisical SoT for host-service shared secrets

2026-06-07 21:04:44 +00:00

legal-court-fetch-drain.config.cjs

feat(X13): scheduled drain — fully-autonomous digest→fetch→ingest loop

2026-06-07 20:31:53 +00:00

legal-court-fetch-service.config.cjs

docs: record Infisical SoT for host-service shared secrets

2026-06-07 21:04:44 +00:00

legal-digest-drain.config.cjs

feat(ops): /operations — מוני-תור אחידים, "מה רץ עכשיו", וניהול-תהליכים

2026-06-08 08:57:23 +00:00

legal-halacha-drain.config.cjs

feat(halacha): drain לילי (23:00–05:00) + per-upload חילוץ תיק-בודד דרך ה-CEO (#120)

2026-06-11 14:02:38 +00:00

legal-metadata-drain.config.cjs

feat(extraction): precedent metadata via Gemini Flash + scheduled drainer

2026-06-08 05:13:49 +00:00

legal-reaper.config.cjs

fix(X13): harden court-fetch against browser leaks + reaper for task-master-mcp leak

2026-06-07 19:43:53 +00:00

migrate_blobs_to_minio.py

feat(storage): #106.4 — DB-driven blob→MinIO migration script (dry-run default)

2026-06-11 17:19:05 +00:00

migrate_gap51_outcomes.py

feat(mcp): FU-14 GAP-51 — איחוד אוצר-המילים של תוצאת-תיק (set_outcome SSoT)

2026-06-06 15:34:49 +00:00

monitor_halacha_quality.py

feat: Stage C — RAG advanced (#33, #47, #48, #49, #50, #51)

2026-05-26 11:26:52 +00:00

multimodal_backfill.py

feat(retrieval): add voyage-multimodal-3 page-image embeddings (feature flag)

2026-05-03 19:24:52 +00:00

nevo_corpus_audit.py

feat(halacha): #86.2 nevo-leak audit + safe ratio backfill · #86.3 ratio-coverage benchmark

2026-06-11 16:50:50 +00:00

nevo_ratio_benchmark.py

feat(nevo): backfill leaked preamble + ratio gold-set benchmark (#86)

2026-06-06 19:45:43 +00:00

notify.py

CEO: add email notifications, subtask parentId, and Paperclip UI assets

2026-04-14 15:55:55 +00:00

pc.sh

feat(paperclip): close 11 integration gaps (#16-#28)

2026-05-04 17:25:45 +00:00

process_pending_blam.py

fix(cases): בל"מ badge reads proceeding_type, not just appeal_subtype

2026-05-26 09:34:23 +00:00

reap_orphan_procs.py

fix(X13): harden court-fetch against browser leaks + reaper for task-master-mcp leak

2026-06-07 19:43:53 +00:00

rechunk_legacy_precedents.py

chore(#57): re-chunk+re-embed legacy precedents (pre-#55 chunker remediation)

2026-06-03 07:55:42 +00:00

reembed_voyage.py

ops: switch embeddings to voyage-3 + plan for context-3 + multimodal-3.5

2026-05-03 16:43:48 +00:00

restore-db.sh

Add full decision writing pipeline: classify, extract, brainstorm, write, QA, export

2026-04-03 10:21:47 +00:00

retrofit_case.py

Add Track Changes architecture for draft revisions (CMP + CMPA)

2026-04-16 18:49:30 +00:00

SCRIPTS.md

feat(storage): #106.4 — DB-driven blob→MinIO migration script (dry-run default)

2026-06-11 17:19:05 +00:00

spec-guard.sh

feat(ci): G12 leak-guard — enforce the Agent Platform Port seam (R4, #113)

2026-06-10 09:40:42 +00:00

style_lesson_panel.py

fix(style-panel): idempotency + dedup — re-running --apply never duplicates lessons

2026-06-08 10:57:57 +00:00

sync_agents_across_companies.py

feat(sync): --verify exits non-zero on drift; adapter mismatch = loud drift (GAP-21, FU-8a)

2026-05-31 11:14:44 +00:00

sync_missing_agent_skills.py

feat(paperclip): close 11 integration gaps (#16-#28)

2026-05-04 17:25:45 +00:00

test_retrieval_by_name.py

fix(retrieval): make decisions findable by name + unhide committee uploads

2026-05-30 11:26:19 +00:00

upload_blam_decisions.py

feat(proceeding-type): explicit ערר/בל"מ field for cases + corpus

2026-05-26 09:17:33 +00:00

voyage_context3_poc_long.py

feat(retrieval): add voyage rerank-2 cross-encoder stage (feature flag)

2026-05-03 18:43:41 +00:00

voyage_context3_poc.py

feat(retrieval): add voyage rerank-2 cross-encoder stage (feature flag)

2026-05-03 18:43:41 +00:00

voyage_multimodal_poc.py

feat(retrieval): add voyage rerank-2 cross-encoder stage (feature flag)

2026-05-03 18:43:41 +00:00

voyage_rerank_corpus_poc.py

feat(retrieval): add voyage rerank-2 cross-encoder stage (feature flag)

2026-05-03 18:43:41 +00:00

voyage_rerank_judge_poc.py

feat(retrieval): add voyage rerank-2 cross-encoder stage (feature flag)

2026-05-03 18:43:41 +00:00