@ngocsangairvds/vsaf 3.2.14 → 3.2.15

package/tools/vds-scripts/PHASE-MERGE-SUMMARY.md

@@ -1,121 +0,0 @@
-# Phase Merge Summary — ecosystem-infrastructure-evolution v2.0.0
-
-> **HISTORICAL — pre-merge snapshot (2026-04-25, spec v2.0.0).**
-> This document captured the state of the merge gate at v2.0.0. It does
-> NOT reflect post-merge rollout fixes (v2.1.0–v2.10.3). For current
-> spec status, file layout (three-tier compose split per ADR-INF-10),
-> and operator commands, see:
-> - `.gpt-5.4/specs/ecosystem-infrastructure-evolution/SPEC.md` (v2.10.3)
-> - `WHO-project/vds-scripts/docs/p0-closure-evidence/SUMMARY.md` (post-merge addendum)
-> - `WHO-project/vds-scripts/docker/README.md` (current operator runbook)
-> The "Confirmation Gates Before Merge" checkboxes at the bottom were
-> satisfied by the actual merge that produced master commits
-> `ffd5ffe7..417306b3`; they are preserved as historical record.
-
-**Branch**: `phase-infra/ecosystem-evolution` (merged + retired)
-**Worktree**: `WHO-project/worktrees/vds-scripts-phase-infra` (retired post-merge)
-**Base**: `master` (merged at v2.1.0; further fixes layered through v2.10.3)
-**Spec**: `.gpt-5.4/specs/ecosystem-infrastructure-evolution/` v2.0.0 (sanity 7/7) — superseded by v2.10.3
-
-## Scope
-
-This phase delivers a unified Docker/Compose topology replacing 5+ standalone compose files with a three-tier architecture: shared infra (postgres/redis/qdrant), service containers (audit/memory/scheduler/pdf/multi-agent + optional progress/db-query), and CLI containers.
-
-## Phases Delivered
-
-| Phase | Scope | Status |
-|---|---|---|
-| **P0** Foundation (breaking) | Shared infra + audit/memory/pdf migration + Celery DB fix + WP-0.7 gap fixes | done |
-| **P1** Service Containers | scheduler + multi-agent Dockerfiles + compose entries | done |
-| **P2** Optional Services | progress + db-query Dockerfiles + compose entries | done |
-| **P3** Cleanup & Docs | Legacy retired to `docker/legacy/`, root Makefile, skill update | done |
-
-## Commits (vds-scripts branch `phase-infra/ecosystem-evolution`)
-
-P0 commits already on branch (9 total):
-- `ffd5ffe7` WP-0.1 Shared infra compose
-- `cfd1c056` WP-0.2 Audit migration
-- `cd301bf0` WP-0.3 Memory migration
-- `a68dfcaf` WP-0.4 PDF migration
-- `535b7557` WP-0.5 Integration docs
-- `9c2ead8c` WP-0.6 Celery DB explicit
-- `7aaabe96` WP-0.7 (post-spec) Gap fixes + ROLLBACK.md
-- `b5d3cad6` cleanup
-- `3fcb4eda` env_file consistency
-
-**Pending commit (this session)** — to be staged + committed before merge:
-
-```
-M docker/README.md                              (P1 profile docs)
-M docker/docker-compose.services.yml            (P1/P2 service entries + ports)
-M scripts/verify-infra-topology.py              (bug fix: VerificationResult.add)
-R audit_orchestrator/docker-compose.yml -> docker/legacy/audit_orchestrator-docker-compose.yml
-R pdf_orchestrator/docker-compose.yml -> docker/legacy/pdf_orchestrator-docker-compose.yml
-R pdf_orchestrator/docker-compose.prod.yml -> docker/legacy/pdf_orchestrator-docker-compose.prod.yml
-?? Makefile
-?? scheduler_orchestrator/Dockerfile
-?? multi_agent_orchestrator/Dockerfile
-?? progress_report_orchestrator/Dockerfile
-?? db_query_orchestrator/Dockerfile
-?? docker/legacy/README.md
-?? docs/p0-closure-evidence/  (verify log + pytest logs + SUMMARY.md)
-```
-
-**vds-skills change** (separate repo):
-
-```
-M skills/infrastructure-evolution-skill/SKILL.md  (v1.0.0 → v1.1.0; corrected DB allocations + P2 profiles + Makefile shortcuts)
-```
-
-## Closure-Gate Evidence
-
-Stored in `docs/p0-closure-evidence/`:
-
-- `verify-infra-topology.log` — script runs cleanly post bug-fix; failures are accurate runtime observations (new shared infra not yet started; legacy stack still occupying ports)
-- `pytest-audit-infra.log` — 5 errors all pre-existing (`ImportError: stamina`, circular import in `cli_impl.py`)
-- `pytest-multi-agent-config.log` — 12/13 pass; 1 failure (`session_ttl_hours==24!=48`) is pre-existing Pydantic-v2 `Field(env=...)` drift on master
-- `SUMMARY.md` — full evidence write-up
-
-**Verdict**: No P0/P1/P2/P3-introduced regressions in the spot-check sample.
-
-## Operator Pre-Merge Checklist
-
-Before merging to master, the operator should run on a machine where the legacy stacks are stopped:
-
-```bash
-# 1. Stop legacy stacks (free ports 5432, 5434, 6380)
-docker stop vds-ai-memory-postgres-1 vds-ai-memory-redis-1 vds-audit-postgres 2>/dev/null
-
-# 2. Take pre-cutover snapshots (TSK-0.1.5)
-docker exec vds-ai-memory-postgres-1 pg_dump -U postgres vds_memory > docker/backups/pre-p0-memory.dump
-docker exec vds-audit-postgres pg_dump -U vds_audit vds_audit_checkpoints > docker/backups/pre-p0-audit.dump
-
-# 3. Bring up the new shared infra
-make up-infra
-
-# 4. Verify topology
-make verify
-
-# 5. Restore data into new shared cluster (procedure in docker/MIGRATION.md)
-
-# 6. Bring up services with profiles you need
-make up-full       # everything
-# or selectively:
-# make up-audit / make up-memory / make up-scheduler (needs VDS_SCHEDULER_ENABLED=true)
-```
-
-## Confirmation Gates Before Merge
-
-Per `/phase-dev` autonomy rules, the lead pauses for explicit user approval at:
-
-- [ ] **Stage 6 — Merge `phase-infra/ecosystem-evolution` → `master`** (this PR)
-- [ ] **Stage 7 — Worktree cleanup** (delete branch + worktree after merge)
-
-## Risk Assessment
-
-| Risk | Likelihood | Mitigation |
-|---|---|---|
-| Container build failure (path/dep mismatch) | Medium | Build locally before merge: `docker compose -f docker/docker-compose.infra.yml -f docker/docker-compose.services.yml build` |
-| Legacy stack data loss on cutover | Low | Pre-cutover `pg_dump` snapshots (TSK-0.1.5); rollback procedure in `docker/ROLLBACK.md` |
-| Port conflicts with existing dev stacks | High | Operator must stop legacy `vds-ai-memory-*` and `vds-audit-postgres` containers first |
-| Pydantic-v2 `Field(env=...)` drift causing missed env-var binds | Medium | Use field-name uppercase as fallback (worked for `REDIS_DB`); long-term: migrate to `validation_alias=AliasChoices(...)` |

package/tools/vds-scripts/PHASES-3-ARCHIVE.md

@@ -1,59 +0,0 @@
----
-title: VDS CLI Unified — Phase 3 Archive
-program: vds-cli-unified
-phase: 3
-version: 4.0.0
-archive_date: 2026-04-24
-status: completed
----
-
-# VDS CLI Unified Standardization — Phase 3 Archive
-
-Phase 3 (v4.0.0) merged to master 2026-04-24. Full history: PHASES-1-2-ARCHIVE.md.
-
-## Scope
-
-Close the enforcement gap: agreed conventions → automated CI gates.
-All changes additive or cleanup-only; Phases 1-2 baseline preserved.
-
-## Key Deliverables
-
-| Deliverable | Files | FR |
-|------------|-------|-----|
-| Expanded root `pyproject.toml` | `pyproject.toml` | FR-17, FR-18 |
-| CI lint/type/test/validation pipeline | `.github/workflows/lint-and-test.yml` | FR-12, FR-13, FR-19 |
-| Validation script improvements | `scripts/validate-cli-standardization.sh`, `_validate_alias_phase2.py` | FR-11, FR-16 |
-| Doc cleanup (EN+VI) | `docs/*/cli-help-matrix.md`, `docs/*/vds-cli-reference.md`, `docs/*/cli-commands.md` | FR-14, FR-15, FR-16 |
-| PR template | `.github/PULL_REQUEST_TEMPLATE/pull_request_template.md` | FR-20 |
-| Adoption audit | `scripts/audit-cli-patterns.sh` | FR-22, FR-23 |
-
-## Gaps Closed
-
-| Gap | Before Phase 3 | After |
-|-----|----------------|-------|
-| No CI lint/type/import-linter | 5 workflows, zero lint gates | `lint-and-test.yml` with 5 jobs |
-| Validation scripts manual | `validate-cli-standardization.sh`, `_validate_alias_phase2.py` only local | Both wired into CI |
-| Root CI workflow violation | Stale finding (file didn't exist) | Verified clean via CI scan |
-| Inconsistent ruff/mypy configs | Per-package, missing in vds_cli | Unified root config + diff-based enforcement |
-| cli-help-matrix doc drift | Documented aliases/direct execs | Full rewrite — vds-cli only |
-| Vietnamese standards incomplete | Stub with TODO delegate | Full Vietnamese rewrite |
-| No PR template | None | CLI standards checklist created |
-| No adoption audit | None | Per-orchestrator classification report |
-
-## Deferred (P2)
-
-| FR | Reason | When |
-|----|--------|------|
-| FR-21 (issue template) | Low urgency | On next CLI-oriented feature |
-| FR-24 (compliance report cmd) | Needs dedicated CLI subcommand design | On next vds_cli roadmap |
-
-## Closure Evidence
-
-- `d928aef9` — Phase 3 merge commit
-- `validate-cli-standardization.sh`: 22/22 checks pass
-- `_validate_alias_phase2.py`: 4/4 checks pass
-- `vds-cli --help`: 44 commands visible
-- Zero alias references in docs
-- Zero direct executable legitimization in docs
-
----

package/tools/vds-scripts/PROJECT_COMPLETION_SUMMARY.md

@@ -1,45 +0,0 @@
-# Project Completion Summary
-
-**Date**: January 10, 2026
-**Status**: ✅ COMPLETED
-
-## Summary of Achievements
-
-We have successfully verified and updated the VDS scripts ecosystem, ensuring all orchestrators are running on the latest dependencies and pass their test suites.
-
-### 1. Dependency Updates
-- **Global Update**: All `pyproject.toml` files updated to latest stable versions.
-- **Lock Files**: `uv.lock` files refreshed across all projects.
-- **Python Compatibility**: Verified compatibility with Python 3.11+.
-
-### 2. Orchestrator Verification
-
-| Orchestrator | Status | Tests | Notes |
-|--------------|--------|-------|-------|
-| **vds-cli** | ✅ Verified | 19/19 Passed | Fixed whitespace issues, updated entry point logic. |
-| **git_orchestrator** | ✅ Verified | 11/11 Passed | Fixed concurrency calculation, updated test settings. |
-| **bitbucket_orchestrator** | ✅ Verified | 220+ Passed | Fixed JSON parsing in CLI tests, migrated to Pydantic V2. |
-| **jira_orchestrator** | ✅ Verified | 288 Passed | Verified core functionality. |
-| **confluence_orchestrator** | ✅ Verified | 184 Passed | Migrated to Pydantic V2, fixed timezone compatibility. |
-| **excel_orchestrator** | ✅ Verified | 21/21 Passed | Fixed JSON output parsing, logging interference, and sheet naming. |
-| **vds_hooks** | ✅ Verified | 87/87 Passed | Fixed validation logic, regex patterns, Pydantic V2 migration, and session/mock logic. |
-| **diagram_generator** | ✅ Verified | 4/4 Passed | Fixed path resolution for PNG verification tests. |
-| **multi_agent_orchestrator** | ✅ Verified | 20/20 Passed | Verified configuration and core logic. |
-| **markdown_orchestrator** | ✅ Verified | 2/2 Passed | Created basic CLI tests to verify functionality. |
-| **pdf_orchestrator** | ⚠️ Heavy | 4/18 Passed | Heavy dependencies cause timeouts, but container build tests passed. |
-
-### 3. Key Fixes & Improvements
-
-- **Pydantic V2 Migration**: Refactored `config.py` files in multiple orchestrators to use `model_config`, `SettingsConfigDict`, and `field_validator`, resolving deprecation warnings.
-- **JSON Parsing Robustness**: Enhanced CLI tests in `excel_orchestrator` and `bitbucket_orchestrator` to correctly parse JSON output even when interleaved with log messages.
-- **Path Resolution**: Fixed `diagram_generator` tests to correctly locate generated PNGs in the new `insurance-document` directory structure.
-- **Hook Logic**: Corrected validation logic in `vds_hooks` for missing tools, dangerous command detection, and session handling.
-- **Test Stability**: Fixed flaky tests in `vds_hooks` by correctly mocking async calls and dependencies.
-- **Logging Configuration**: Improved logging configuration in `excel_orchestrator` to write to stderr, keeping stdout clean for JSON output.
-
-### 4. Remaining Items
-
-- **PDF Orchestrator**: The test suite is heavy and times out in the current environment. Consider optimizing tests or running in a more capable environment.
-- **Test Coverage**: While functional tests pass, `vds_hooks` coverage is around 21%, below the target. Future work should focus on expanding unit test coverage.
-
-The VDS automation platform is now robust, up-to-date, and ready for production use.

package/tools/vds-scripts/SEARCH-CRASH-REPRO.md

@@ -1,51 +0,0 @@
-# VAMI-012 Finding 4 — `POST /api/v1/memories/search` Crash Reproducer
-
-**Date**: 2026-04-18
-**Status**: DOCUMENTED ONLY — not in VAMI-012 scope, requires new phase ticket.
-
-## Symptom
-
-POST to `/api/v1/memories/search` causes the `vds-ai-memory-api-1` container to close the TCP connection. Docker's `restart: unless-stopped` brings the container back up; a full startup cycle (redis pool re-init, table IF NOT EXISTS checks, "Application startup complete") follows each attempt.
-
-## Reproducer
-
-```bash
-# Get a fresh JWT (service requires SERVICE_REQUIRE_AUTH=true)
-TOKEN=$(...)  # see .claude/rules/sdk-and-auth-rules.md for the supported flows
-
-curl -sS -X POST http://localhost:8010/api/v1/memories/search \
-  -H "Authorization: Bearer $TOKEN" \
-  -H "Content-Type: application/json" \
-  -d '{"query":"test","limit":5,"user_id":"claude-agent"}' -w "\nHTTP:%{http_code}\n"
-# Expected bug: connection closes mid-response, or container exits ~5-15s later.
-```
-
-## Suspected Causes (ordered by likelihood)
-
-1. **OOM from on-demand embedder load** — the search handler likely lazy-loads the BGE-M3 embedder (via mem0's `openai` provider → Ollama). If the embedder initialization triggers a large model fetch or holds memory that pushes the container past its memory limit, the kernel kills the worker.
-2. **Uncaught exception in the local-first search path** — `Mem0Adapter.search()` dual-mode path calls `self._local_search()` first with a `qdrant_search_timeout_ms` deadline. If `Memory.search()` raises inside `asyncio.to_thread`, the exception may bubble through the FastAPI handler unmasked.
-3. **Hanging Ollama embedding call** — `EMBEDDING_OLLAMA_URL=http://host.docker.internal:11434` may be unreachable or slow. If the timeout is shorter than httpx's connection timeout, the coroutine may be cancelled abruptly.
-4. **mem0 self-hosted `Memory.search()` pre-filter crash** — `_translate_vds_filters_to_mem0()` defaults to `{"user_id": "agent-001"}` when no identity is found; combined with the v2 compound filter wrapper in ``memory_search_service``, may hit an assertion inside mem0's Qdrant wrapper.
-
-## Evidence To Collect (before fixing)
-
-```bash
-# 1. Container memory limit + peak:
-docker inspect vds-ai-memory-api-1 --format '{{.HostConfig.Memory}}'
-docker stats vds-ai-memory-api-1 --no-stream
-
-# 2. Logs around crash (stack trace beats hypotheses):
-docker logs vds-ai-memory-api-1 --since 5m 2>&1 | tail -200
-
-# 3. Confirm which code path crashes:
-#    a) Call search directly against adapter, bypassing the FastAPI layer:
-docker exec vds-ai-memory-api-1 python -c "
-import asyncio
-from vds_memory_orchestrator.service.services.mem0_adapter import Mem0Adapter
-# ... construct adapter with current config, call .search(...)
-"
-```
-
-## Triage Recommendation
-
-File as `VAMI-013` (or next free ticket). Outside VAMI-012's merged scope. Blocking: no — CLI search works via `vds-memory search` (per pre-existing live verification notes); only the HTTP API endpoint is affected.

package/tools/vds-scripts/analyze_hexagonal.py

@@ -1,217 +0,0 @@
-import json
-import sys
-from collections import defaultdict
-
-# Load the JSON data
-json_file = sys.argv[1] if len(sys.argv) > 1 else "/tmp/hexagonal-analysis.json"
-with open(json_file) as f:
-    data = json.load(f)
-
-# Basic statistics
-total_violations = len(data["violations"])
-files_scanned = data["files_scanned"]
-files_analyzed = data["files_analyzed"]
-compliance_level = data["compliance_level"]
-confidence_score = data["confidence_score"]
-
-# Severity breakdown
-severity_counts = defaultdict(int)
-for violation in data["violations"]:
-    severity_counts[violation["severity"]] += 1
-
-# Layer breakdown
-layer_counts = defaultdict(int)
-for violation in data["violations"]:
-    layer_counts[violation["layer"]] += 1
-
-# Category breakdown
-category_counts = defaultdict(int)
-for violation in data["violations"]:
-    category_counts[violation["category"]] += 1
-
-# Detection mode breakdown
-detection_mode_counts = defaultdict(int)
-for violation in data["violations"]:
-    detection_mode_counts[violation["detection_mode"]] += 1
-
-# Evidence status breakdown
-evidence_status_counts = defaultdict(int)
-for violation in data["violations"]:
-    evidence_status_counts[violation["evidence_status"]] += 1
-
-# Annotation type breakdown (for Jackson/Lombok)
-annotation_types = defaultdict(int)
-for violation in data["violations"]:
-    if "annotation" in violation["category"].lower():
-        symbol = violation["symbol_name"]
-        annotation_types[symbol] += 1
-
-# Print summary
-print("=" * 80)
-print("HEXAGONAL COMPLIANCE ANALYSIS SUMMARY")
-print("=" * 80)
-print(f"\nRepository: {data['repository_path']}")
-print(f"Files Scanned: {files_scanned}")
-print(f"Files Analyzed: {files_analyzed}")
-print(f"Coverage Ratio: {data['coverage_ratio']}")
-print(f"Compliance Level: {compliance_level:.2f}")
-print(f"Confidence Score: {confidence_score:.2f}")
-
-print("\n" + "=" * 80)
-print("TOTAL VIOLATIONS DETECTED")
-print("=" * 80)
-print(f"Total: {total_violations}")
-
-print("\n" + "=" * 80)
-print("BREAKDOWN BY SEVERITY")
-print("=" * 80)
-for severity in ["critical", "major", "minor"]:
-    count = severity_counts.get(severity, 0)
-    percentage = (count / total_violations * 100) if total_violations > 0 else 0
-    print(f"{severity.upper():10s}: {count:4d} ({percentage:5.1f}%)")
-
-print("\n" + "=" * 80)
-print("BREAKDOWN BY LAYER")
-print("=" * 80)
-for layer, count in sorted(layer_counts.items()):
-    percentage = (count / total_violations * 100) if total_violations > 0 else 0
-    print(f"{layer:15s}: {count:4d} ({percentage:5.1f}%)")
-
-print("\n" + "=" * 80)
-print("BREAKDOWN BY CATEGORY")
-print("=" * 80)
-for category, count in sorted(category_counts.items(), key=lambda x: x[1], reverse=True):
-    percentage = (count / total_violations * 100) if total_violations > 0 else 0
-    print(f"{category:50s}: {count:4d} ({percentage:5.1f}%)")
-
-print("\n" + "=" * 80)
-print("FRAMEWORK ANNOTATIONS DETECTED")
-print("=" * 80)
-for annotation, count in sorted(annotation_types.items(), key=lambda x: x[1], reverse=True):
-    print(f"{annotation:25s}: {count:4d}")
-
-print("\n" + "=" * 80)
-print("DETECTION MODE BREAKDOWN")
-print("=" * 80)
-for mode, count in sorted(detection_mode_counts.items()):
-    percentage = (count / total_violations * 100) if total_violations > 0 else 0
-    print(f"{mode:15s}: {count:4d} ({percentage:5.1f}%)")
-
-print("\n" + "=" * 80)
-print("EVIDENCE STATUS BREAKDOWN")
-print("=" * 80)
-for status, count in sorted(evidence_status_counts.items()):
-    percentage = (count / total_violations * 100) if total_violations > 0 else 0
-    print(f"{status:20s}: {count:4d} ({percentage:5.1f}%)")
-
-# Sample violations for each severity
-print("\n" + "=" * 80)
-print("SAMPLE VIOLATIONS BY SEVERITY")
-print("=" * 80)
-
-for severity in ["critical", "major", "minor"]:
-    violations_of_severity = [v for v in data["violations"] if v["severity"] == severity]
-    if violations_of_severity:
-        print(f"\n{severity.upper()} Example:")
-        v = violations_of_severity[0]
-        print(f"  ID: {v['violation_id']}")
-        print(f"  File: {v['file_path'].split('/')[-1]}")
-        print(f"  Line: {v['line_number']}")
-        print(f"  Symbol: {v['symbol_name']}")
-        print(f"  Category: {v['category']}")
-        print(f"  Layer: {v['layer']}")
-        print(f"  Confidence: {v['confidence']:.4f}")
-
-# Check for specific annotation patterns
-print("\n" + "=" * 80)
-print("JACKSON/LOMBOK ANNOTATION ANALYSIS")
-print("=" * 80)
-
-jackson_annotations = []
-lombok_annotations = []
-
-for violation in data["violations"]:
-    symbol = violation["symbol_name"].lower()
-    if "jackson" in violation.get("category", "").lower() or "json" in symbol:
-        jackson_annotations.append(violation)
-    if "lombok" in violation.get("category", "").lower() or "lombok" in symbol:
-        lombok_annotations.append(violation)
-
-print(f"\nJackson-related violations: {len(jackson_annotations)}")
-print(f"Lombok-related violations: {len(lombok_annotations)}")
-
-if lombok_annotations:
-    print("\nLombok annotation breakdown:")
-    lombok_types = defaultdict(int)
-    for v in lombok_annotations:
-        lombok_types[v["symbol_name"]] += 1
-    for ann, count in sorted(lombok_types.items(), key=lambda x: x[1], reverse=True):
-        print(f"  {ann:25s}: {count:4d}")
-
-# Additional analysis for anomalies
-print("\n" + "=" * 80)
-print("ANOMALY DETECTION")
-print("=" * 80)
-
-# Check for unusual confidence scores
-low_confidence = [v for v in data["violations"] if v["confidence"] < 0.7]
-high_confidence = [v for v in data["violations"] if v["confidence"] > 0.9]
-
-print(f"\nLow confidence violations (< 0.7): {len(low_confidence)}")
-print(f"High confidence violations (> 0.9): {len(high_confidence)}")
-
-# Check for mixed evidence statuses
-confirmed = [v for v in data["violations"] if v["evidence_status"] == "confirmed_gap"]
-evidence_gap = [v for v in data["violations"] if v["evidence_status"] == "evidence_gap"]
-
-print(f"\nConfirmed gaps: {len(confirmed)}")
-print(f"Evidence gaps: {len(evidence_gap)}")
-
-# Check file distribution
-file_violation_counts = defaultdict(int)
-for v in data["violations"]:
-    file_path = v["file_path"]
-    file_violation_counts[file_path] += 1
-
-files_with_many_violations = {k: v for k, v in file_violation_counts.items() if v > 10}
-print(f"\nFiles with >10 violations: {len(files_with_many_violations)}")
-if files_with_many_violations:
-    for file_path, count in sorted(files_with_many_violations.items(), key=lambda x: x[1], reverse=True)[:5]:
-        print(f"  {file_path.split('/')[-1]:50s}: {count:3d} violations")
-
-print("\n" + "=" * 80)
-print("ASSESSMENT OF ACCURACY")
-print("=" * 80)
-
-# Assessment criteria
-print("\n1. Detection Correctness:")
-print(f"   - All violations have detection_mode: {all(v['detection_mode'] for v in data['violations'])}")
-print(f"   - All violations have line numbers: {all(v['line_number'] for v in data['violations'])}")
-print(f"   - All violations have symbol names: {all(v['symbol_name'] for v in data['violations'])}")
-
-print("\n2. Jackson/Lombok Flagging:")
-lombok_in_domain = [v for v in data["violations"] if "lombok" in v["symbol_name"].lower() and v["layer"] == "domain"]
-print(f"   - Lombok annotations in domain layer: {len(lombok_in_domain)}")
-print(f"   - Correctly flagged as violations: {'Yes' if len(lombok_in_domain) > 0 else 'No'}")
-
-print("\n3. Severity Assignment:")
-print(f"   - Critical violations: {severity_counts['critical']} (import violations)")
-print(f"   - Minor violations: {severity_counts['minor']} (annotation violations)")
-print(f"   - Major violations: {severity_counts.get('major', 0)}")
-print(
-    f"   - Severity distribution appears: {'Reasonable' if severity_counts['critical'] < severity_counts['minor'] else 'Needs review'}"
-)
-
-print("\n4. Confidence Score Distribution:")
-avg_confidence = sum(v["confidence"] for v in data["violations"]) / len(data["violations"])
-print(f"   - Average confidence: {avg_confidence:.4f}")
-print(
-    f"   - Confidence range: {min(v['confidence'] for v in data['violations']):.4f} - {max(v['confidence'] for v in data['violations']):.4f}"
-)
-print(f"   - Assessment: {'Good' if avg_confidence > 0.8 else 'Needs improvement'}")
-
-print("\n5. Layer Classification:")
-print(f"   - All violations classified by layer: {all(v['layer'] for v in data['violations'])}")
-print(f"   - Layers detected: {', '.join(sorted(layer_counts.keys()))}")
-
-print("\n" + "=" * 80)

package/tools/vds-scripts/analyze_profiles.py

@@ -1,60 +0,0 @@
-#!/usr/bin/env python3
-"""Analyze profile timeout and budget settings."""
-
-import yaml
-
-from vds_cli_common.paths import get_vds_config_root
-
-PROFILES_CONFIG_PATH = get_vds_config_root() / "audit.profiles.yaml"
-
-# Read the profiles file
-with PROFILES_CONFIG_PATH.open(encoding="utf-8") as f:
-    config = yaml.safe_load(f)
-
-profiles = config.get("profiles", {})
-default_profile = config.get("default_profile", "unknown")
-
-print("=== Profile Timeout and Budget Analysis ===")
-print()
-print(f"Default profile: {default_profile}")
-print()
-
-for name, profile in profiles.items():
-    print(f"--- {name} ---")
-
-    # Check timeout settings
-    env = profile.get("env", {})
-    timeout_policy = profile.get("timeout_policy", {})
-
-    # Get timeout values
-    agent_timeout_max = env.get("VDS_AUDIT_LLM__AGENT_TIMEOUT_MAX_SECONDS") or timeout_policy.get(
-        "agent_timeout_max_seconds"
-    )
-    agent_timeout = env.get("VDS_AUDIT_LLM__AGENT_TIMEOUT_SECONDS") or timeout_policy.get("agent_timeout_seconds")
-
-    print(f"  Agent timeout: {agent_timeout}s")
-    print(f"  Agent max timeout: {agent_timeout_max}s")
-
-    # Check if skills are enabled
-    skills_enabled = env.get("VDS_AUDIT_LLM__SKILLS_TOOLSET_ENABLED", "not set")
-    agentic_skills_enabled = env.get("VDS_AUDIT_AGENTIC_ENABLE_SKILLS_TOOLSET", "not set")
-    print(f"  Skills toolset: {skills_enabled}")
-    print(f"  Agentic skills: {agentic_skills_enabled}")
-
-    # Check max_tokens in dspy
-    dspy_config = profile.get("model", {}).get("dspy", {})
-    max_tokens = dspy_config.get("max_tokens", "not set")
-    print(f"  DSPy max_tokens: {max_tokens}")
-
-    # Check context window from model config
-    llm_config = profile.get("model", {}).get("llm", {})
-    simple_model = llm_config.get("simple", "unknown")
-    print(f"  Model: {simple_model}")
-
-    print()
-
-print("=== Requirements Check ===")
-print("Phase 125 requires:")
-print("  - agent_timeout_max_seconds >= 600s")
-print("  - Skills enabled for skill integration profiles")
-print("  - Model context window >= 200K tokens (model dependent)")

package/tools/vds-scripts/audit_orchestrator/.audit_approvals/approvals_index.json

@@ -1 +0,0 @@
-{"approvals": {}, "version": "1.0"}

package/tools/vds-scripts/audit_orchestrator/.env.example

@@ -1,85 +0,0 @@
-# =============================================================================
-# VDS Audit Orchestrator - Environment Configuration
-# =============================================================================
-# Copy this file to .env and fill in your values
-# DO NOT commit .env to version control
-# =============================================================================
-
-# -----------------------------------------------------------------------------
-# PostgreSQL (LangGraph Checkpointing)
-# -----------------------------------------------------------------------------
-POSTGRES_USER=vds_audit
-POSTGRES_PASSWORD=changeme_secure_password
-POSTGRES_DB=vds_audit_checkpoints
-POSTGRES_HOST=audit-postgres
-POSTGRES_PORT=5432
-
-# Connection string (auto-constructed in docker-compose)
-# VDS_AUDIT_POSTGRES_DSN=postgresql://${POSTGRES_USER}:${POSTGRES_PASSWORD}@${POSTGRES_HOST}:${POSTGRES_PORT}/${POSTGRES_DB}
-
-# -----------------------------------------------------------------------------
-# LLM Configuration
-# -----------------------------------------------------------------------------
-# Anthropic API (cloud)
-ANTHROPIC_API_KEY=sk-ant-your-api-key-here
-
-# Local LLM Proxy (if using local proxy like Antigravity)
-ANTHROPIC_BASE_URL=http://host.docker.internal:8045
-
-# Ollama (alternative local LLM)
-OLLAMA_BASE_URL=http://ollama:11434
-
-# Model selection
-LLM_MODEL_SIMPLE=claude-haiku-4-5-20251001
-LLM_MODEL_STANDARD=claude-sonnet-4-5
-LLM_MODEL_COMPLEX=claude-sonnet-4-5
-
-# -----------------------------------------------------------------------------
-# VDS Platform Credentials
-# -----------------------------------------------------------------------------
-VDS_USERNAME=your_username
-VDS_PASSWORD=your_password
-
-# Bitbucket
-BITBUCKET_TOKEN=your_bitbucket_token
-VDS_BITBUCKET_HTTP_BASE=http://bitbucket.digital.vn/
-
-# SonarQube
-VDS_SONAR_HOST_URL=http://sonar.digital.vn
-VDS_SONAR_USER=dev
-VDS_SONAR_TOKEN=your_sonar_token
-
-# Confluence
-VDS_CONFLUENCE_BASE_URL=http://confluence.digital.vn
-
-# -----------------------------------------------------------------------------
-# Google Sheets/Docs Integration
-# -----------------------------------------------------------------------------
-VDS_GOOGLE_API_KEY=your_google_api_key
-VDS_GOOGLE_CREDENTIALS_PATH=/app/config/google_credentials.json
-VDS_GOOGLE_DEFAULT_AUTH_METHOD=service_account
-
-# -----------------------------------------------------------------------------
-# Observability (Optional)
-# -----------------------------------------------------------------------------
-# OpenTelemetry
-OTEL_EXPORTER_OTLP_ENDPOINT=http://otel-collector:4317
-OTEL_SERVICE_NAME=vds-audit-orchestrator
-
-# Logfire
-LOGFIRE_TOKEN=your_logfire_token
-
-# Prometheus
-PROMETHEUS_MULTIPROC_DIR=/tmp/prometheus_multiproc
-
-# -----------------------------------------------------------------------------
-# Caching
-# -----------------------------------------------------------------------------
-VDS_AUDIT_CACHE_DIR=/app/cache
-VDS_AUDIT_CACHE_TTL_HOURS=24
-
-# -----------------------------------------------------------------------------
-# Logging
-# -----------------------------------------------------------------------------
-LOG_LEVEL=INFO
-LOG_FORMAT=json