refactor: remove obsolete V1 implementation paths

.gitignore

@@ -1,8 +1,4 @@
-backend/.venv/
 node_modules/
-frontend/.next/
-frontend/.next-dev-log/
-frontend/node_modules/
 **/.next/
 **/.next-dev-log/
 **/node_modules/
@@ -14,7 +10,6 @@ __pycache__/
 *.egg-info/
 .DS_Store
 ._*
-backend/.pytest_cache/
 
 # Local secrets and generated artifacts
 .env.deploytmp
@@ -22,12 +17,8 @@ backend/.pytest_cache/
 **/.env.development.local
 qa-artifacts/
 storage/
-backend/.env
-backend/storage/
-backend/tmp_migration_check.db
 project-docs/qa-artifacts/
 project-docs/automation-artifacts/
-backend/project-docs/qa-artifacts/
 .claude/worktrees/
 .gstack/
 .runtime/
@@ -38,8 +29,6 @@ pip-audit-project.json
 .codex/phase1-mainline-signoff/
 .codex-temp/
 output/
-frontend/next-env.d.ts
-frontend/tsconfig.tsbuildinfo
 **/next-env.d.ts
 **/tsconfig.tsbuildinfo
 

README.md

@@ -1,8 +1,6 @@
 # Yuqei AI Contract Platform V2
 
-This repository `main` branch is now the clean V2 rebuild line.
-
-The previous complete V1 implementation is preserved on the `V1` branch. New product, architecture, database, AI platform, frontend, backend, deployment, and QA work should be developed on `main` using the final V2 four-project layout from day one.
+This repository contains only the active V2 product line. Product, architecture, database, AI platform, frontend, backend, deployment, and QA work use the final four-project layout below.
 
 ## Project Layout
 
@@ -37,14 +35,12 @@ docs/
 project-docs/
 ```
 
-## Branch Policy
+## Repository Policy
 
-- `V1`: frozen snapshot of the original system before the V2 rebuild.
-- `main`: active V2 rebuild development branch.
-
-Do not continue adding new product features to the V1 application tree on `main`. If V1 code is needed for reference, check the `V1` branch.
-
-Do not reintroduce the old root `apps/`、`services/`、`packages/`、`sdks/`、`infra/` layout. V2 is organized around final product boundaries: contract platform, AI platform, Python SDK, and infrastructure.
+- `main` is the active V2 development line.
+- Do not reintroduce a retired product tree or parallel product implementation.
+- Keep the final product boundaries: Contract Platform, AI Platform, Python SDK, and infrastructure.
+- Protocol identifiers such as `/api/v1`, persisted schema versions, migration history, and source-provenance labels are compatibility contracts inside V2; they are not a second product implementation.
 
 ## Starting Points
 

V2-README.md

@@ -1,8 +1,6 @@
 # Yuqei AI Contract Platform V2
 
-This branch is the clean V2 rebuild line.
-
-The previous V1 implementation is preserved on the `V1` branch. New V2 development starts from the final four-project layout below instead of continuing to modify the V1 application tree.
+This repository contains only the active V2 product line. Development follows the final four-project layout below.
 
 ## Target Projects
 
@@ -42,5 +40,5 @@ docs/
 
 ## Branch Policy
 
-- `V1`: frozen snapshot of the current pre-rebuild system.
-- `main`: V2 rebuild development line.
+- `main`: active V2 development line.
+- Retired product trees and parallel product implementations must not be reintroduced.

project-docs/V2-next-development-optimization-plan-2026-06-28.md

@@ -51,7 +51,7 @@
 
 任务：
 
-1. 使用 `yuqei-infra/scripts/inspect-v2-database-state.ps1` 只读检查远端数据库状态。
+1. 只读检查远端数据库状态（当时的一次性盘点工具已在 V2-only 收口中退役；当前按现行运行核验流程执行）。
 2. 启动并核验 Docker smoke 环境 `yuqei-contract-v2-smoke-*`。
 3. 确认 smoke PostgreSQL 中是否存在：
    - `yuqei_contract`

project-docs/V2-release-readiness-checklist-2026-07-10.md

@@ -19,7 +19,7 @@ V2 当前候选版本已完成本地回归和 `192.168.3.90` PostgreSQL smoke
 | 核心页面 smoke | 通过 | 工作台、模板起草、合同台账、模板库、合同仓库、数据看板均返回 200 |
 | 核心 API smoke | 通过 | contracts/templates/workspace/files 均返回 200 |
 | 前端类型检查 | 通过 | `npm run typecheck` 通过 |
-| 资产计数脚本 | 通过 | `yuqei-infra/scripts/compare-v1-v2-assets.py` 已生成 artifact |
+| 资产计数脚本 | 历史通过 | 当时的对比工具已生成 artifact；该一次性工具已在 V2-only 收口中退役 |
 | 远端 Contract API schema-check | 通过 | PostgreSQL smoke 返回 `ok` |
 | 远端 AI Platform schema-check | 通过 | PostgreSQL smoke 返回 `ok`，pgvector 维度与索引通过 |
 | 远端 Docker smoke | 通过 | Web、API、worker、PostgreSQL、Redis、MinIO 均健康 |
@@ -45,17 +45,7 @@ Artifacts：
 | V1 Prompt 黄金样本文件 | 3 | - | 已作为来源参考 |
 | V2 AI golden asset 文件 | - | 89 | 已迁入 AI 平台资产目录 |
 
-重复执行命令：
-
-```powershell
-python yuqei-infra\scripts\compare-v1-v2-assets.py --contract-api-base http://localhost:8101
-```
-
-远端 smoke 资产计数命令：
-
-```powershell
-python yuqei-infra\scripts\compare-v1-v2-assets.py --contract-api-base http://192.168.3.90:18100
-```
+当时使用的 V1/V2 一次性对比命令已退役；以上表格和 artifact 仅保留为 2026-07-10 历史证据。当前资产完整性由现行 V2 资产门禁验证。
 
 ## Release Candidate 冻结
 

project-docs/V2-vs-V1-feature-gap-audit-2026-07-03.md

@@ -1,82 +0,0 @@
-# AI 合同平台 V2 对照 V1 功能缺口审计
-
-日期：2026-07-03
-
-## 对照来源
-
-- V1 线上入口：`http://contact.yuqei.com/`
-- V1 源码目录：`D:\yuqei-project\ai-contract-platform-v1`
-- V2 源码目录：`D:\yuqei-project\ai-contract-platform`
-
-本次以源码为准，线上 V1 入口用于确认公开产品形态。
-
-## 总体结论
-
-V2 已经吸收并强化了 V1 的核心合同能力，包括：
-
-- 模板库、变量 schema、模板树、模板发布/下线/回滚/历史。
-- 标准条款库、审查规则库、条款推荐、规则下线边界。
-- 法律全文库精选迁移、法律版本表、法规变更影响扫描。
-- AI Trace、质量动作、Prompt/模型/知识库管理。
-- 审查结果结构化、风险项/引用依据只读 API、条款修复。
-- 合同审批、版本、文件、通知、履约、签署证据、归档、统一搜索。
-- Docker smoke、schema-check、接口压测、安全扫描、CI 质量门。
-
-当前仍值得从 V1 回补的主要不是“合同核心骨架”，而是产品化和交付体验层能力。
-
-## 建议加入但 V2 尚未完整闭环
-
-| 优先级 | V1 能力 | V2 当前状态 | 是否建议加入 | 建议方式 |
-| --- | --- | --- | --- | --- |
-| P0 | 审查报告 PDF/报告交付 | V2 有审查风险、引用依据和文件系统，但未看到审查报告生成/下载 API 与前端入口 | 强烈建议 | 基于 V2 结构化审查结果生成报告，不迁 V1 旧实现；先做 HTML/PDF 模板、报告文件入库、下载事件审计 |
-| P1 | 用户、租户设置、账号管理 | V2 有 tenant context 和轻量访问控制，但没有完整用户/租户后台页面 | 建议加入 | 做最小管理员后台：用户列表、角色、租户配置、默认配额；先不做复杂组织架构 |
-| P1 | 全局审计日志查看 | V2 有合同时间线、审计完整性和 AI Trace，但缺全局 Contract audit log 页面 | 建议加入 | 增加管理员审计中心，聚合合同事件、下载事件、审批、模板/法律资产发布事件 |
-| P1 | 法律文书起草体系 | V1 有 `legal-document-types`、element schema、draft、generate-from-draft 和移动端入口；V2 主要是合同模板起草 | 建议分阶段加入 | 先迁“催告函、解除通知、授权委托书、法律意见书”这类高频非合同文书，沿用 V2 模板/快照/版本化模型 |
-| P1 | 用户侧法律问答/Chat 页面 | V2 后端已有 `/api/v1/legal/qa` 代理和 AI Platform legal QA，但前端没有独立用户问答页 | 建议加入 | 做轻量法律问答页，支持来源卡、Trace 链接、从合同风险项带上下文进入问答 |
-| P2 | 公开首页、演示申请、试点申请 | V1 有 landing、pricing、demo-request、pilot-requests；V2 首页直接进入工作台 | 若 V2 要对外发布则建议 | 做公开首页 + 演示申请，先不做完整订单支付；申请数据进入销售/试点队列 |
-| P2 | 商业套餐、计费、订单/发票 | V1 有 pricing、billing、commercial plans、order/invoice 模型 | 可后置 | 先迁套餐展示和用量概览；真实支付、发票、订单后台等商业闭环等产品上线后再做 |
-| P2 | 销售跟进和试点运营 | V1 有 sales-followups、pilot readiness、operation-kpis | 可后置但有价值 | 若要做商业化试点，加入试点看板、跟进事件、试点 KPI；内部交付可暂缓 |
-| P2 | 运营 KPI 看板 | V2 有 dashboard 和质量门文档，但没有 V1 `operation-kpis` 运营指标页 | 建议后置 | 复用 V2 审查、报告、模板、Trace、知识库指标，做运营看板 |
-| P3 | 移动端 Flutter 客户端 | V1 mobile 已覆盖登录、审查、报告、合同生成、法律文书、Chat、版本链、真机 smoke | 暂不建议现在加入 | V2 Web 收口后单独立项；不要把 V1 Flutter 代码直接塞进 V2 主线 |
-| P3 | 形式化上线治理工作台 | V1 有 governance assets/imports/releases/workbench | 可按企业交付需要加入 | V2 已有 schema governance 和质量门；若做正式客户交付，可做“交付证据/签收/上线门禁”管理页 |
-
-## 已加入或基本覆盖，不应重复开发
-
-| V1 能力 | V2 状态 |
-| --- | --- |
-| 模板化合同起草 | 已覆盖，V2 还有发布快照和生命周期管理 |
-| 30 个 Phase13 合同模板与变量 schema | 已覆盖，V1 当前 Phase13 模板数和 V2 迁移资产均为 30 |
-| 标准条款库和审查规则库 | 已覆盖，V2 有管理页、diff、校验、发布快照 |
-| 法律全文库精选与法规影响扫描 | 已覆盖，V2 已有 `legal_fulltext_versions` |
-| 黄金样本与质量评估资产 | 已覆盖，V2 AI golden gate 已纳入 Docker smoke |
-| Prompt/模型/知识库/Trace 管理 | 已覆盖，V2 AI 管理后台更完整 |
-| 文件仓储、审批、合同版本、通知、履约、签署证据、归档 | V2 已覆盖 |
-| V1 同步 auto-review 体验 | 不建议迁，V2 异步 worker 更合理 |
-| V1 单体后端结构 | 不建议迁，V2 拆分架构更合理 |
-
-## 不建议直接迁移
-
-- V1 单体 FastAPI 后端结构。
-- V1 页面大而全布局。
-- V1 同步审查调用体验。
-- V1 旧 PDF 实现本身。应保留“报告交付”产品目标，用 V2 结构化审查结果重建。
-- V1 商业订单/发票/支付后台的完整实现，除非 V2 确认进入对外商用阶段。
-
-## 建议下一步
-
-最值得先做的是 P0：审查报告 PDF/报告交付。
-
-原因：
-
-1. V1 线上产品叙事明确包含“报告交付”。
-2. V2 已有结构化风险、引用依据、条款推荐、文件仓储和下载审计，生成报告的基础数据已经齐。
-3. 报告是客户可感知的交付物，比继续扩后台运营更能补齐产品闭环。
-
-建议第一片范围：
-
-- `GET /api/v1/contracts/{contract_id}/review-report`
-- `POST /api/v1/contracts/{contract_id}/review-report`
-- `GET /api/v1/contracts/{contract_id}/review-report/download`
-- 报告内容包含合同摘要、风险分布、风险项、引用依据、条款推荐、修复状态、免责声明。
-- 报告文件进入 V2 file repository，下载写入审计事件。
-- 前端在合同详情页审查区增加“生成/下载审查报告”入口。

yuqei-ai-platform/packages/ai-openapi/openapi.yaml

@@ -1063,7 +1063,7 @@ paths:
   /api/v1/quality/golden-samples:
     get:
       operationId: getGoldenSampleCoverage
-      summary: Get V1 golden sample coverage imported into the V2 AI quality loop
+      summary: Get packaged golden sample coverage used by the V2 AI quality loop
       responses:
         "200":
           description: Golden sample coverage
@@ -1074,7 +1074,7 @@ paths:
   /api/v1/quality/golden-samples/evaluate:
     post:
       operationId: evaluateGoldenSamples
-      summary: Run asset integrity evaluation for V1 golden samples
+      summary: Run asset integrity evaluation for packaged golden samples
       responses:
         "200":
           description: Golden sample evaluation report

yuqei-ai-platform/services/ai-platform-api/README.md

@@ -27,7 +27,7 @@ python -m alembic upgrade head
 
 ## Golden Sample Asset Gate
 
-The V1 Phase9, Phase13, and Phase15 golden sample assets are bundled with this
+The packaged Phase9, Phase13, and Phase15 golden sample assets are bundled with this
 service and can be checked without starting FastAPI:
 
 ```powershell

yuqei-ai-platform/services/ai-platform-api/pyproject.toml

@@ -23,7 +23,7 @@ dependencies = [
 [project.scripts]
 yuqei-golden-sample-gate = "yuqei_ai_platform_api.golden_sample_gate:main"
 yuqei-release-readiness-gate = "yuqei_ai_platform_api.release_readiness_gate:main"
-yuqei-import-v1-prompt-assets = "yuqei_ai_platform_api.v1_prompt_assets_import:main"
+yuqei-import-packaged-prompt-assets = "yuqei_ai_platform_api.packaged_prompt_assets_import:main"
 
 [build-system]
 requires = ["hatchling"]

yuqei-ai-platform/services/ai-platform-api/src/yuqei_ai_platform_api/golden_sample_gate.py

@@ -14,7 +14,7 @@ from yuqei_ai_platform_api.golden_samples import (
 
 def main(argv: Sequence[str] | None = None) -> int:
     parser = argparse.ArgumentParser(
-        description="Run the V1 golden sample asset integrity gate for the V2 AI platform.",
+        description="Run the packaged golden sample asset integrity gate for the V2 AI platform.",
     )
     parser.add_argument(
         "--format",

yuqei-ai-platform/services/ai-platform-api/src/yuqei_ai_platform_api/golden_samples.py

@@ -42,7 +42,7 @@ class GoldenSampleSuiteRequirement(BaseModel):
 class GoldenSampleCoverage(BaseModel):
     schema_version: str
     source_product: str = "ai-contract-platform-v1"
-    source_description: str = "V1 Phase9 quality fixtures and Phase15 prompt golden samples."
+    source_description: str = "Packaged quality fixtures and prompt golden samples used by V2."
     generated_at: datetime
     total_assets: int
     total_loaded_samples: int
@@ -94,34 +94,34 @@ class GoldenSampleEvaluationReport(BaseModel):
 PHASE9_CSV_ASSETS = [
     {
         "suite_id": "phase9_contract_review",
-        "title": "V1 Phase9 合同审查黄金样本",
+        "title": "内置 Phase9 合同审查黄金样本",
         "source_file": "golden-review-samples.csv",
         "prompt_type": "contract_review",
         "sample_family": "phase9_review",
     },
     {
         "suite_id": "phase9_legal_chat",
-        "title": "V1 Phase9 知识库问答黄金样本",
+        "title": "内置 Phase9 知识库问答黄金样本",
         "source_file": "golden-kb-question-set.csv",
         "prompt_type": "legal_chat",
         "sample_family": "phase9_kb_qa",
     },
     {
         "suite_id": "phase9_contract_generation",
-        "title": "V1 Phase9 合同生成黄金样本",
+        "title": "内置 Phase9 合同生成黄金样本",
         "source_file": "golden-contract-generation-set.csv",
         "prompt_type": "contract_generation",
         "sample_family": "phase9_generation",
     },
     {
         "suite_id": "phase9_quality_metrics",
-        "title": "V1 Phase9 质量指标字典",
+        "title": "内置 Phase9 质量指标字典",
         "source_file": "quality-metric-dictionary.csv",
         "sample_family": "phase9_quality_metrics",
     },
     {
         "suite_id": "phase9_prompt_versions",
-        "title": "V1 Phase9 Prompt 版本登记",
+        "title": "内置 Phase9 Prompt 版本登记",
         "source_file": "prompt-version-register.csv",
         "sample_family": "phase9_prompt_governance",
     },
@@ -130,14 +130,14 @@ PHASE9_CSV_ASSETS = [
 PHASE15_JSON_ASSETS = [
     {
         "suite_id": "phase15_wording_rewrite",
-        "title": "V1 Phase15 条款润色黄金样本",
+        "title": "内置 Phase15 条款润色黄金样本",
         "source_file": "wording-rewrite-golden-samples.json",
         "prompt_type": "wording_rewrite",
         "sample_family": "phase15_wording_rewrite",
     },
     {
         "suite_id": "phase15_business_auxiliary",
-        "title": "V1 Phase15 业务辅助黄金样本",
+        "title": "内置 Phase15 业务辅助黄金样本",
         "source_file": "business-auxiliary-golden-samples.json",
         "prompt_type": "business_auxiliary",
         "sample_family": "phase15_business_auxiliary",
@@ -147,21 +147,21 @@ PHASE15_JSON_ASSETS = [
 PHASE13_CSV_ASSETS = [
     {
         "suite_id": "phase13_contract_generation",
-        "title": "V1 Phase13 模板化合同生成黄金样本",
+        "title": "内置 Phase13 模板化合同生成黄金样本",
         "source_file": "inventories/generation-golden-sample-inventory.csv",
         "prompt_type": "contract_generation",
         "sample_family": "phase13_generation",
     },
     {
         "suite_id": "phase13_contract_review",
-        "title": "V1 Phase13 合同审查黄金样本",
+        "title": "内置 Phase13 合同审查黄金样本",
         "source_file": "inventories/review-golden-sample-inventory.csv",
         "prompt_type": "contract_review",
         "sample_family": "phase13_review",
     },
     {
         "suite_id": "phase13_legal_chat",
-        "title": "V1 Phase13 法律问答黄金样本",
+        "title": "内置 Phase13 法律问答黄金样本",
         "source_file": "inventories/kb-qa-golden-sample-inventory.csv",
         "prompt_type": "legal_chat",
         "sample_family": "phase13_kb_qa",
@@ -171,7 +171,7 @@ PHASE13_CSV_ASSETS = [
 PHASE13_MARKDOWN_SET_ASSETS = [
     {
         "suite_id": "phase13_clause_recommendation",
-        "title": "V1 Phase13 条款推荐黄金样本",
+        "title": "内置 Phase13 条款推荐黄金样本",
         "source_file": "qa-samples/generation/GEN-SAMPLE-CLAUSE-*-expected.md",
         "prompt_type": "clause_recommendation",
         "sample_family": "phase13_clause_generation",
@@ -198,8 +198,8 @@ def load_golden_sample_coverage() -> GoldenSampleCoverage:
     return GoldenSampleCoverage(
         schema_version=str(manifest.get("schema_version", "v1-golden-samples")),
         source_description=(
-            "V1 Phase9 quality fixtures, Phase13 legal knowledge template library, "
-            "and Phase15 prompt golden samples."
+            "Packaged Phase9 quality fixtures, Phase13 legal knowledge template library, "
+            "and Phase15 prompt golden samples used by V2."
         ),
         generated_at=datetime.now(UTC),
         total_assets=len(assets),
@@ -223,7 +223,7 @@ def run_golden_sample_evaluation() -> GoldenSampleEvaluationReport:
             observed=coverage.ready_required_suite_count,
             detail=(
                 f"{coverage.ready_required_suite_count}/{coverage.required_suite_count} "
-                "个 V1 manifest 必需套件已达到最低样本数。"
+                "个资产 manifest 必需套件已达到最低样本数。"
             ),
         ),
         _count_check(
@@ -231,7 +231,7 @@ def run_golden_sample_evaluation() -> GoldenSampleEvaluationReport:
             title="黄金样本总量",
             expected=coverage.total_loaded_samples,
             observed=coverage.total_loaded_samples,
-            detail=f"已加载 {coverage.total_loaded_samples} 条 V1 黄金样本资产。",
+            detail=f"已加载 {coverage.total_loaded_samples} 条内置黄金样本资产。",
         ),
     ]
     empty_assets = [asset.suite_id for asset in coverage.assets if asset.status != "ready" or asset.loaded_samples <= 0]
@@ -406,10 +406,10 @@ def _requirement_to_coverage(
     loaded_samples = _loaded_samples_for_requirement(prompt_type, sample_family, assets)
     if loaded_samples >= min_samples:
         status: Literal["ready", "below_minimum", "not_loaded"] = "ready"
-        coverage_note = "样本数量达到 V1 manifest 要求。"
+        coverage_note = "样本数量达到资产 manifest 要求。"
     elif loaded_samples > 0:
         status = "below_minimum"
-        coverage_note = f"已接入 {loaded_samples} 条，低于 V1 manifest 要求的 {min_samples} 条。"
+        coverage_note = f"已接入 {loaded_samples} 条，低于资产 manifest 要求的 {min_samples} 条。"
     else:
         status = "not_loaded"
         coverage_note = "当前 V2 只读资产包尚未接入该套件样本。"

yuqei-ai-platform/services/ai-platform-api/src/yuqei_ai_platform_api/packaged_prompt_assets_import.py

@@ -17,7 +17,7 @@ from yuqei_ai_platform_api.repository import build_sqlalchemy_engine
 
 
 DEFAULT_VERSION = "v1-phase15"
-DEFAULT_CHANGED_BY = "v1-prompt-assets-import"
+DEFAULT_CHANGED_BY = "packaged-prompt-assets-import"
 DEFAULT_SOURCE_FILE = (
     Path(__file__).resolve().parent
     / "prompt_assets"
@@ -27,7 +27,7 @@ DEFAULT_SOURCE_FILE = (
 
 
 @dataclass(frozen=True)
-class V1PromptAsset:
+class PackagedPromptAsset:
     prompt_key: str
     prompt_type: str
     name: str
@@ -52,7 +52,7 @@ class PromptImportResult:
 
     def as_payload(self) -> dict[str, Any]:
         return {
-            "command": "import-v1-prompt-assets",
+            "command": "import-packaged-prompt-assets",
             "mode": self.mode,
             "source_file": self.source_file,
             "source_schema_version": self.source_schema_version,
@@ -68,16 +68,16 @@ class PromptImportResult:
         }
 
 
-def load_v1_prompt_assets(source_file: Path) -> tuple[list[V1PromptAsset], str | None, list[str]]:
+def load_packaged_prompt_assets(source_file: Path) -> tuple[list[PackagedPromptAsset], str | None, list[str]]:
     payload = json.loads(source_file.read_text(encoding="utf-8-sig"))
     if not isinstance(payload, dict):
-        raise ValueError("V1 prompt asset file must contain a JSON object.")
+        raise ValueError("Packaged prompt asset file must contain a JSON object.")
     schema_version = _optional_str(payload.get("schema_version"))
     assets_payload = payload.get("assets")
     if not isinstance(assets_payload, list):
-        raise ValueError("V1 prompt asset file must contain an assets array.")
+        raise ValueError("Packaged prompt asset file must contain an assets array.")
 
-    assets_by_key: dict[str, V1PromptAsset] = {}
+    assets_by_key: dict[str, PackagedPromptAsset] = {}
     duplicate_keys: list[str] = []
     for index, item in enumerate(assets_payload):
         if not isinstance(item, dict):
@@ -85,7 +85,7 @@ def load_v1_prompt_assets(source_file: Path) -> tuple[list[V1PromptAsset], str |
         prompt_key = _required_str(item.get("prompt_key"), f"assets[{index}].prompt_key")
         prompt_type = _optional_str(item.get("prompt_type")) or "unknown"
         name = _optional_str(item.get("name")) or prompt_key
-        asset = V1PromptAsset(
+        asset = PackagedPromptAsset(
             prompt_key=prompt_key,
             prompt_type=prompt_type,
             name=name,
@@ -98,9 +98,9 @@ def load_v1_prompt_assets(source_file: Path) -> tuple[list[V1PromptAsset], str |
     return list(assets_by_key.values()), schema_version, sorted(set(duplicate_keys))
 
 
-def import_v1_prompt_assets(
+def import_packaged_prompt_assets(
     session_factory,
-    assets: list[V1PromptAsset],
+    assets: list[PackagedPromptAsset],
     *,
     source_file: Path,
     source_schema_version: str | None,
@@ -202,7 +202,7 @@ def import_v1_prompt_assets(
                             before_template_json=_json_dumps(before_payload) if before_payload is not None else None,
                             after_template_json=_json_dumps(after_payload),
                             changed_by=changed_by,
-                            change_note=change_note or "Import V1 Phase15 default prompt asset.",
+                            change_note=change_note or "Import packaged default prompt asset.",
                             created_at=now,
                         )
                     )
@@ -224,12 +224,12 @@ def import_v1_prompt_assets(
 
 
 def build_parser() -> argparse.ArgumentParser:
-    parser = argparse.ArgumentParser(description="Import V1 Phase15 default prompt assets into V2 prompt_templates.")
+    parser = argparse.ArgumentParser(description="Import packaged default prompt assets into V2 prompt_templates.")
     parser.add_argument(
         "--source-file",
         type=Path,
         default=DEFAULT_SOURCE_FILE,
-        help="Prompt asset JSON path. Defaults to the packaged V1 Phase15 asset file.",
+        help="Prompt asset JSON path. Defaults to the packaged prompt asset file.",
     )
     parser.add_argument("--database-url", default=None, help="Override AI_PLATFORM_DATABASE_URL.")
     parser.add_argument("--version", default=DEFAULT_VERSION, help="Prompt template version to upsert.")
@@ -247,8 +247,8 @@ def main() -> None:
     engine = build_sqlalchemy_engine(database_url)
     session_factory = sessionmaker(bind=engine, expire_on_commit=False, future=True)
     try:
-        assets, schema_version, duplicate_keys = load_v1_prompt_assets(args.source_file)
-        result = import_v1_prompt_assets(
+        assets, schema_version, duplicate_keys = load_packaged_prompt_assets(args.source_file)
+        result = import_packaged_prompt_assets(
             session_factory,
             assets,
             source_file=args.source_file,

yuqei-ai-platform/services/ai-platform-api/src/yuqei_ai_platform_api/prompts.py

@@ -55,7 +55,7 @@ REVIEW_PROMPT_DOMAIN_KEY_BY_CONTRACT_TYPE = {
     "data_processing": "contract_review_tech_saas_data_compliance",
 }
 
-_V1_PROMPT_TOKEN = re.compile(r"{{\s*([A-Za-z0-9_]+)\s*}}")
+_PACKAGED_PROMPT_TOKEN = re.compile(r"{{\s*([A-Za-z0-9_]+)\s*}}")
 
 
 def legal_qa_prompt_candidates() -> tuple[str, ...]:
@@ -82,7 +82,7 @@ def render_legal_qa_prompt(
         "legal_article_refs": _format_citations(citations),
     }
     if prompt_template:
-        imported_prompt = _render_v1_prompt_asset(prompt_template.content, variables)
+        imported_prompt = _render_packaged_prompt_asset(prompt_template.content, variables)
         if imported_prompt:
             return imported_prompt
     template = prompt_template.content if prompt_template else _default_legal_qa_template()
@@ -108,7 +108,7 @@ def render_contract_review_prompt(
         "structured_output_instructions": structured_output_instructions,
     }
     if prompt_template:
-        imported_prompt = _render_v1_prompt_asset(
+        imported_prompt = _render_packaged_prompt_asset(
             prompt_template.content,
             variables,
             supplemental_context=(
@@ -126,7 +126,7 @@ def render_contract_review_prompt(
     return f"{rendered}\n\n{structured_output_instructions}"
 
 
-def _render_v1_prompt_asset(
+def _render_packaged_prompt_asset(
     content: str,
     variables: dict[str, str],
     *,
@@ -145,7 +145,7 @@ def _render_v1_prompt_asset(
     if not prompt_content:
         return None
     system_prompt = _optional_text(asset.get("system_prompt"))
-    rendered_content = _V1_PROMPT_TOKEN.sub(
+    rendered_content = _PACKAGED_PROMPT_TOKEN.sub(
         lambda match: variables.get(match.group(1), "未提供"),
         prompt_content,
     )

yuqei-ai-platform/services/ai-platform-api/tests/test_packaged_prompt_assets_import.py

@@ -5,14 +5,14 @@ from sqlalchemy import create_engine, select
 from sqlalchemy.orm import sessionmaker
 
 from yuqei_ai_platform_api.models import Base, PromptTemplateChangeModel, PromptTemplateModel
-from yuqei_ai_platform_api.v1_prompt_assets_import import (
+from yuqei_ai_platform_api.packaged_prompt_assets_import import (
     DEFAULT_SOURCE_FILE,
-    import_v1_prompt_assets,
-    load_v1_prompt_assets,
+    import_packaged_prompt_assets,
+    load_packaged_prompt_assets,
 )
 
 
-def test_load_v1_prompt_assets_dedupes_by_prompt_key(tmp_path: Path) -> None:
+def test_load_packaged_prompt_assets_dedupes_by_prompt_key(tmp_path: Path) -> None:
     source_file = tmp_path / "default-prompt-assets.json"
     source_file.write_text(
         json.dumps(
@@ -44,7 +44,7 @@ def test_load_v1_prompt_assets_dedupes_by_prompt_key(tmp_path: Path) -> None:
         encoding="utf-8",
     )
 
-    assets, schema_version, duplicate_keys = load_v1_prompt_assets(source_file)
+    assets, schema_version, duplicate_keys = load_packaged_prompt_assets(source_file)
 
     assert schema_version == "phase15-default-prompt-assets.v1"
     assert duplicate_keys == ["legal_chat_default"]
@@ -54,8 +54,8 @@ def test_load_v1_prompt_assets_dedupes_by_prompt_key(tmp_path: Path) -> None:
     assert rendered["asset"]["prompt_content"] == "second"
 
 
-def test_packaged_v1_prompt_asset_file_is_complete() -> None:
-    assets, schema_version, duplicate_keys = load_v1_prompt_assets(DEFAULT_SOURCE_FILE)
+def test_packaged_prompt_asset_file_is_complete() -> None:
+    assets, schema_version, duplicate_keys = load_packaged_prompt_assets(DEFAULT_SOURCE_FILE)
 
     assert DEFAULT_SOURCE_FILE.exists()
     assert schema_version == "phase15-default-prompt-assets.v1"
@@ -71,7 +71,7 @@ def test_packaged_v1_prompt_asset_file_is_complete() -> None:
     }
 
 
-def test_import_v1_prompt_assets_is_idempotent_and_publishes(tmp_path: Path) -> None:
+def test_import_packaged_prompt_assets_is_idempotent_and_publishes(tmp_path: Path) -> None:
     source_file = tmp_path / "default-prompt-assets.json"
     source_file.write_text(
         json.dumps(
@@ -89,12 +89,12 @@ def test_import_v1_prompt_assets_is_idempotent_and_publishes(tmp_path: Path) ->
         ),
         encoding="utf-8",
     )
-    assets, schema_version, duplicate_keys = load_v1_prompt_assets(source_file)
+    assets, schema_version, duplicate_keys = load_packaged_prompt_assets(source_file)
     engine = create_engine(f"sqlite+pysqlite:///{(tmp_path / 'prompts.sqlite').as_posix()}", future=True)
     Base.metadata.create_all(engine)
     session_factory = sessionmaker(bind=engine, expire_on_commit=False, future=True)
 
-    first = import_v1_prompt_assets(
+    first = import_packaged_prompt_assets(
         session_factory,
         assets,
         source_file=source_file,
@@ -102,7 +102,7 @@ def test_import_v1_prompt_assets_is_idempotent_and_publishes(tmp_path: Path) ->
         apply=True,
         duplicate_prompt_keys=duplicate_keys,
     )
-    second = import_v1_prompt_assets(
+    second = import_packaged_prompt_assets(
         session_factory,
         assets,
         source_file=source_file,

yuqei-ai-platform/services/ai-platform-api/tests/test_platform_data.py

@@ -1012,7 +1012,7 @@ def test_quality_rules_can_be_hot_updated_and_drive_audit_labels() -> None:
     assert audit["quality_score"] == 90
 
 
-def test_golden_sample_coverage_loads_v1_quality_assets() -> None:
+def test_golden_sample_coverage_loads_packaged_quality_assets() -> None:
     client = make_client()
 
     response = client.get("/api/v1/quality/golden-samples")

yuqei-ai-platform/services/ai-platform-api/tests/test_provider_flow.py

@@ -510,14 +510,14 @@ def test_render_legal_qa_prompt_includes_question_context_and_citations() -> Non
     assert "第七百零三条" in prompt
 
 
-def test_legal_qa_prefers_published_v1_prompt_asset() -> None:
+def test_legal_qa_prefers_published_packaged_prompt_asset() -> None:
     repository = InMemoryAiPlatformRepository()
     repository.seed_defaults()
     created = repository.upsert_prompt_template(
         PromptTemplateCreate(
             name="legal_chat_default",
             version="v1-phase15",
-            content=_v1_prompt_content(
+            content=_packaged_prompt_content(
                 system_prompt="V1 法律问答专项系统提示。",
                 prompt_content=(
                     "问题={{question}}；背景={{review_context}}；依据={{legal_article_refs}}。"
@@ -552,7 +552,7 @@ def test_contract_review_selects_and_renders_domain_prompt_asset() -> None:
         PromptTemplateCreate(
             name="contract_review_investment_finance",
             version="v1-phase15",
-            content=_v1_prompt_content(
+            content=_packaged_prompt_content(
                 system_prompt="V1 投融资专项审查提示。",
                 prompt_content=(
                     "类型={{contract_type}}；立场={{party_position}}；"
@@ -588,7 +588,7 @@ def test_contract_review_selects_and_renders_domain_prompt_asset() -> None:
     assert "{{" not in audit["prompt_preview"]
 
 
-def test_contract_review_prompt_candidates_cover_v1_business_domains() -> None:
+def test_contract_review_prompt_candidates_cover_packaged_business_domains() -> None:
     expected = {
         "investment-agreement": "contract_review_investment_finance",
         "shareholder-resolution": "contract_review_corporate_governance",
@@ -603,13 +603,13 @@ def test_contract_review_prompt_candidates_cover_v1_business_domains() -> None:
         assert candidates[-2:] == ("contract_review_default", "contract_review.default")
 
 
-def test_render_contract_review_prompt_supports_v1_asset_json() -> None:
+def test_render_contract_review_prompt_supports_packaged_asset_json() -> None:
     repository = InMemoryAiPlatformRepository()
     created = repository.upsert_prompt_template(
         PromptTemplateCreate(
             name="contract_review_default",
             version="v1-phase15",
-            content=_v1_prompt_content(
+            content=_packaged_prompt_content(
                 system_prompt="审慎审查并保留依据。",
                 prompt_content="类型={{contract_type}}；正文={{contract_text}}；依据={{legal_article_refs}}。",
             ),
@@ -633,7 +633,7 @@ def test_render_contract_review_prompt_supports_v1_asset_json() -> None:
     assert "{{" not in prompt
 
 
-def _v1_prompt_content(*, system_prompt: str, prompt_content: str) -> str:
+def _packaged_prompt_content(*, system_prompt: str, prompt_content: str) -> str:
     return json.dumps(
         {
             "source": {

yuqei-contract-platform/apps/contract-web/src/app/admin/ai-prompts/page.tsx

@@ -16,13 +16,13 @@ type PromptTemplate = {
   updated_at: string;
 };
 
-type V1PromptVariable = {
+type ImportedPromptVariable = {
   name?: string;
   required?: boolean;
   description?: string;
 };
 
-type V1PromptAsset = {
+type ImportedPromptAsset = {
   prompt_key?: string;
   prompt_type?: string;
   name?: string;
@@ -30,7 +30,7 @@ type V1PromptAsset = {
   target_scene?: string;
   system_prompt?: string;
   prompt_content?: string;
-  variables?: V1PromptVariable[];
+  variables?: ImportedPromptVariable[];
   output_schema?: unknown;
   model_config?: Record<string, unknown>;
   change_summary?: string;
@@ -42,9 +42,9 @@ type ParsedPromptContent = {
     phase?: string;
     schema_version?: string | null;
   };
-  asset?: V1PromptAsset;
+  asset?: ImportedPromptAsset;
   raw: string;
-  isV1Asset: boolean;
+  isPackagedAsset: boolean;
 };
 
 type LoadState =
@@ -156,7 +156,7 @@ export default function AiPromptsPage() {
           templates.find((template) => options.preferTemplateId && template.id === options.preferTemplateId) ??
           templates.find((template) => options.preferName && template.name === options.preferName && template.status === "published") ??
           templates.find((template) => promptTemplateMatchesTarget(template, currentTarget)) ??
-          templates.find((template) => template.status === "published" && template.version === "v1-phase15") ??
+          templates.find((template) => template.status === "published" && template.enabled) ??
           templates.find((template) => template.status === "published") ??
           templates[0];
         setDraftFromTemplate(nextTemplate);
@@ -229,7 +229,7 @@ export default function AiPromptsPage() {
     .filter((template) => template.name === selectedName)
     .sort((left, right) => compareDate(right.published_at ?? right.updated_at, left.published_at ?? left.updated_at));
   const publishedTemplate = selectedVersions.find((template) => template.status === "published");
-  const v1ImportedCount = templates.filter((template) => template.version === "v1-phase15").length;
+  const packagedAssetCount = templates.filter((template) => parsedById.get(template.id)?.isPackagedAsset).length;
   const publishedCount = templates.filter((template) => template.status === "published").length;
 
   useEffect(() => {
@@ -309,7 +309,7 @@ export default function AiPromptsPage() {
       if (!response.ok) {
         throw new Error(payload?.message ?? `HTTP ${response.status}`);
       }
-      setMessage(`已发布 ${template.name} / ${template.version}。`);
+      setMessage(`已发布 ${template.name} / ${formatVersionLabel(template.version)}。`);
       await loadTemplates({ clearMessage: false, preferTemplateId: payload.id, preferName: template.name });
     } catch (error) {
       setMessage(error instanceof Error ? error.message : "Prompt 发布失败");
@@ -331,7 +331,7 @@ export default function AiPromptsPage() {
       if (!response.ok) {
         throw new Error(payload?.message ?? `HTTP ${response.status}`);
       }
-      setMessage(`已从 ${template.version} 回滚并发布新版本。`);
+      setMessage(`已从 ${formatVersionLabel(template.version)} 回滚并发布新版本。`);
       await loadTemplates({ clearMessage: false, preferTemplateId: payload.id, preferName: template.name });
     } catch (error) {
       setMessage(error instanceof Error ? error.message : "Prompt 回滚失败");
@@ -345,7 +345,7 @@ export default function AiPromptsPage() {
       <div className="page-heading">
         <div>
           <h1>Prompt 中心</h1>
-          <p>管理 AI 中台提示词草稿、发布版本、回滚记录和 V1 Phase15 导入资产。</p>
+          <p>管理 AI 中台提示词草稿、发布版本、回滚记录和平台内置资产。</p>
         </div>
         <div className="toolbar toolbar-inline">
           <button className="button button-soft" onClick={() => void loadTemplates()}>
@@ -361,7 +361,7 @@ export default function AiPromptsPage() {
 
       <section className="template-library-summary" aria-label="Prompt 资产统计">
         <Metric label="Prompt 版本" value={templates.length} />
-        <Metric label="V1 Phase15" value={v1ImportedCount} />
+        <Metric label="内置资产" value={packagedAssetCount} />
         <Metric label="已发布" value={publishedCount} />
         <Metric label="Prompt 类型" value={promptTypes.length} />
       </section>
@@ -403,7 +403,7 @@ export default function AiPromptsPage() {
                 <select className="search-input select-input" value={versionFilter} onChange={(event) => setVersionFilter(event.target.value)}>
                   <option value="">全部版本</option>
                   {versions.map((version) => (
-                    <option key={version} value={version}>{version}</option>
+                    <option key={version} value={version}>{formatVersionLabel(version)}</option>
                   ))}
                 </select>
               </label>
@@ -441,7 +441,7 @@ export default function AiPromptsPage() {
                       </div>
                       <div className="template-library-card-meta">
                         <span>{parsed.asset?.prompt_type ?? "未标注类型"}</span>
-                        <span>{template.version}</span>
+                        <span>{formatVersionLabel(template.version)}</span>
                         <span>{formatDate(template.updated_at)}</span>
                       </div>
                     </button>
@@ -459,9 +459,9 @@ export default function AiPromptsPage() {
                 <FileText size={16} />
                 详情与编辑
               </h2>
-              {publishedTemplate && <p className="item-meta">当前运行版本：{publishedTemplate.version}</p>}
+              {publishedTemplate && <p className="item-meta">当前运行版本：{formatVersionLabel(publishedTemplate.version)}</p>}
             </div>
-            {selectedParsed?.isV1Asset && <span className="tag tag-primary">V1 Phase15</span>}
+            {selectedParsed?.isPackagedAsset && <span className="tag tag-primary">内置资产</span>}
           </div>
           <div className="panel-body">
             {selectedTemplate && selectedParsed && (
@@ -525,10 +525,10 @@ function PromptMetadata({ parsed, template }: { parsed: ParsedPromptContent; tem
   return (
     <div className="legal-article-meta-grid prompt-metadata-grid">
       <MiniBlock title="模板 ID" value={template.id} />
-      <MiniBlock title="版本" value={template.version} />
+      <MiniBlock title="版本" value={formatVersionLabel(template.version)} />
       <MiniBlock title="状态" value={statusLabels[template.status] ?? template.status} />
       <MiniBlock title="类型" value={asset?.prompt_type ?? "未标注"} />
-      <MiniBlock title="来源阶段" value={parsed.source?.phase ?? "V2"} />
+      <MiniBlock title="来源阶段" value={parsed.isPackagedAsset ? "内置基线" : parsed.source?.phase ?? "V2"} />
       <MiniBlock title="目标场景" value={asset?.target_scene ?? "未标注"} />
     </div>
   );
@@ -617,7 +617,7 @@ function VersionHistory({
           <article className={`quality-action-item ${isTargetPrompt ? "config-targeted" : ""}`} key={template.id}>
             <div className="list-row">
               <div className="list-main">
-                <strong>{template.version}</strong>
+                <strong>{formatVersionLabel(template.version)}</strong>
                 <span className="item-meta">
                   更新：{formatDate(template.updated_at)}
                   {template.published_at ? ` / 发布：${formatDate(template.published_at)}` : ""}
@@ -629,7 +629,7 @@ function VersionHistory({
               {isTargetPrompt && <span className="tag tag-primary target-marker">Target</span>}
             </div>
             {template.rollback_from_version && (
-              <span className="item-meta">回滚来源：{template.rollback_from_version}</span>
+              <span className="item-meta">回滚来源：{formatVersionLabel(template.rollback_from_version)}</span>
             )}
             {showTargetPublish && (
               <div className="inline-message">
@@ -714,17 +714,17 @@ function MiniBlock({ title, value }: { title: string; value: string }) {
 
 function parsePromptContent(content: string): ParsedPromptContent {
   try {
-    const payload = JSON.parse(content) as { source?: ParsedPromptContent["source"]; asset?: V1PromptAsset };
+    const payload = JSON.parse(content) as { source?: ParsedPromptContent["source"]; asset?: ImportedPromptAsset };
     return {
       source: payload.source,
       asset: payload.asset,
       raw: content,
-      isV1Asset: payload.source?.phase === "phase15-prompt-assets" || Boolean(payload.asset?.prompt_key)
+      isPackagedAsset: payload.source?.phase === "phase15-prompt-assets" || Boolean(payload.asset?.prompt_key)
     };
   } catch {
     return {
       raw: content,
-      isV1Asset: false
+      isPackagedAsset: false
     };
   }
 }
@@ -733,6 +733,10 @@ function compareDate(left: string, right: string) {
   return new Date(left).getTime() - new Date(right).getTime();
 }
 
+function formatVersionLabel(version: string) {
+  return version === "v1-phase15" ? "内置基线" : version;
+}
+
 function formatDate(value: string) {
   const date = new Date(value);
   if (Number.isNaN(date.getTime())) {

yuqei-contract-platform/apps/contract-web/src/app/admin/ai-quality-dashboard/page.tsx

@@ -987,7 +987,7 @@ function GoldenSampleCoveragePanel({
       <div className="panel-body">
         <div className="quality-action-stats">
           <span>资产 {coverage.total_assets}</span>
-          <span>来源 {coverage.source_product}</span>
+          <span>来源 平台内置质量样本</span>
           <span>合同生成 {requiredSamplesByPromptType["contract_generation"] ?? 0}</span>
           <span>合同审查 {requiredSamplesByPromptType["contract_review"] ?? 0}</span>
           <span>法律问答 {requiredSamplesByPromptType["legal_chat"] ?? 0}</span>

yuqei-contract-platform/apps/contract-web/src/app/admin/legal-assets/page.tsx

@@ -39,7 +39,7 @@ type PageProps = {
 
 const text = {
   title: "条款规则库",
-  subtitle: "集中查看 V1 迁移到 V2 的标准条款、审查规则、法源绑定和条款修复关联。",
+  subtitle: "集中查看平台内置的标准条款、审查规则、法源绑定和条款修复关联。",
   keywordPlaceholder: "搜索条款、规则、法源、建议内容",
   all: "全部",
   clauses: "标准条款",
@@ -67,7 +67,7 @@ const text = {
   diff: "差异",
   emptyClauses: "没有找到匹配的标准条款。",
   emptyRules: "没有找到匹配的审查规则。",
-  readOnlyNote: "当前资产正文来自 V1 Phase13 静态库，生命周期状态已进入 V2 持久化管理；编辑、diff、发布前校验可作为下一步继续接入。",
+  readOnlyNote: "当前内置资产已进入 V2 持久化生命周期管理；编辑、diff 与发布前校验通过现行资产接口执行。",
 };
 
 const contractTypeLabels: Record<string, string> = {
@@ -248,7 +248,7 @@ export default async function LegalAssetsPage({ searchParams }: PageProps) {
                   </h2>
                   <p className="item-meta">{clauses.length} / {allClauses.length} 条资产匹配当前条件</p>
                 </div>
-                <span className="tag tag-primary">V1 Phase13</span>
+                <span className="tag tag-primary">内置资产</span>
               </div>
               <div className="panel-body legal-asset-list">
                 {clauses.length === 0 && <div className="empty-state empty-state-compact">{text.emptyClauses}</div>}
@@ -274,7 +274,7 @@ export default async function LegalAssetsPage({ searchParams }: PageProps) {
                   </h2>
                   <p className="item-meta">{rules.length} / {allRules.length} 条资产匹配当前条件</p>
                 </div>
-                <span className="tag tag-primary">V1 Phase13</span>
+                <span className="tag tag-primary">内置资产</span>
               </div>
               <div className="panel-body legal-asset-list">
                 {rules.length === 0 && <div className="empty-state empty-state-compact">{text.emptyRules}</div>}
@@ -325,7 +325,7 @@ async function ClauseCard({
         {clause.review_status && <span>{clause.review_status}</span>}
         <span>{clause.legal_sources.length} 法源</span>
         <span>{linkedRuleCount} 规则</span>
-        <span>{clause.lifecycle_version || "v1"}</span>
+        <span>{clause.lifecycle_version || "未版本化"}</span>
         <span>{clause.lifecycle_event_count} 次变更</span>
       </div>
       {history.length > 0 && <LegalAssetHistory events={history} />}
@@ -400,7 +400,7 @@ async function RuleCard({
         {rule.status && <span>{rule.status}</span>}
         {rule.knowledge_card_id && <span>{rule.knowledge_card_id}</span>}
         <span>{rule.legal_sources.length} 法源</span>
-        <span>{rule.lifecycle_version || "v1"}</span>
+        <span>{rule.lifecycle_version || "未版本化"}</span>
         <span>{rule.lifecycle_event_count} 次变更</span>
       </div>
       {history.length > 0 && <LegalAssetHistory events={history} />}

yuqei-contract-platform/apps/contract-web/src/app/admin/legal-fulltext/page.tsx

@@ -12,8 +12,6 @@ import {
 import { ImpactScanPanel } from "./impact-scan-panel";
 
 const PAGE_SIZE = 20;
-const FULL_V1_LEGAL_TOTAL = 789_853;
-const FULL_V1_LEGAL_SOURCE_TOTAL = 19_712;
 
 type PageProps = {
   searchParams?: Promise<{
@@ -56,7 +54,7 @@ export default async function LegalFulltextPage({ searchParams }: PageProps) {
       <div className="page-heading">
         <div>
           <h1>法律全文库</h1>
-          <p>V1 Phase14 法律全文资产已接入 V2，可按关键词、法源、效力状态分页查看，并用于法规变更影响扫描。</p>
+          <p>平台法律全文资产可按关键词、法源、效力状态分页查看，并用于法规变更影响扫描。</p>
         </div>
         <Link className="button button-soft" href="/admin/legal-assets">
           <Scale size={16} />
@@ -65,10 +63,10 @@ export default async function LegalFulltextPage({ searchParams }: PageProps) {
       </div>
 
       <section className="template-library-summary" aria-label="法律全文库统计">
-        <Metric icon={BookOpenText} label="V1 全量条文" value={articlePage.total || FULL_V1_LEGAL_TOTAL} />
-        <Metric icon={Landmark} label="V1 法源规模" value={FULL_V1_LEGAL_SOURCE_TOTAL} />
+        <Metric icon={BookOpenText} label="当前命中条文" value={articlePage.total} />
+        <Metric icon={Landmark} label="本页法源" value={Object.keys(articlePage.source_summary).length} />
         <Metric icon={ShieldCheck} label="本页有效条文" value={effectiveCountOnPage} />
-        <Metric icon={Database} label="当前命中" value={articlePage.total} />
+        <Metric icon={Database} label="本页条文" value={articlePage.items.length} />
       </section>
 
       <form className="toolbar legal-fulltext-toolbar" action="/admin/legal-fulltext">
@@ -103,8 +101,8 @@ export default async function LegalFulltextPage({ searchParams }: PageProps) {
           <div className="panel-body legal-asset-filter-list">
             <SourceFilter sourceIds={matchedSources} activeSourceId={sourceId} />
             <div className="legal-asset-block">
-              <strong>全量库状态</strong>
-              <p>已接入 V1 全量法律库 {formatInteger(FULL_V1_LEGAL_TOTAL)} 条，覆盖约 {formatInteger(FULL_V1_LEGAL_SOURCE_TOTAL)} 个法源。页面按服务端分页读取，不一次性渲染全库。</p>
+              <strong>法律库状态</strong>
+              <p>当前查询由服务端返回 {formatInteger(articlePage.total)} 条；页面按服务端分页读取，不一次性渲染全库。</p>
             </div>
           </div>
         </aside>
@@ -121,7 +119,7 @@ export default async function LegalFulltextPage({ searchParams }: PageProps) {
                   共 {formatInteger(articlePage.total)} 条，当前第 {articlePage.page} 页，每页 {articlePage.page_size} 条
                 </p>
               </div>
-              <span className="tag tag-primary">V1 Phase14</span>
+              <span className="tag tag-primary">平台法律库</span>
             </div>
             <div className="panel-body legal-article-list">
               {articlePage.items.length === 0 && <div className="empty-state empty-state-compact">没有找到匹配的法条。</div>}

yuqei-contract-platform/apps/contract-web/src/app/globals.css

@@ -2949,7 +2949,7 @@ textarea {
   gap: 16px;
 }
 
-/* V1-like visual language, V2-like structure */
+/* Product visual language and current application structure */
 
 .attribution-grid {
   display: grid;
@@ -4626,7 +4626,7 @@ details.panel summary::-webkit-details-marker {
   display: none;
 }
 
-/* V1 visual polish on top of the V2 layout */
+/* Product visual polish for the current layout */
 .panel,
 .metric,
 .metric-card,

yuqei-contract-platform/apps/contract-web/src/app/templates/page.tsx

@@ -29,7 +29,7 @@ type PageProps = {
 
 const text = {
   title: "模板库",
-  subtitle: "按合同类型组织 V1 模板资产，在 V2 中串起起草、审查、条款修复和版本留痕。",
+  subtitle: "按合同类型组织平台内置模板资产，串起起草、审查、条款修复和版本留痕。",
   searchPlaceholder: "搜索模板名称、适用范围、校验规则",
   search: "搜索",
   clear: "清空",

yuqei-contract-platform/services/contract-api/pyproject.toml

@@ -19,7 +19,6 @@ dependencies = [
 ]
 
 [project.scripts]
-yuqei-import-v1-legal-fulltext = "yuqei_contract_api.v1_legal_fulltext_import:main"
 yuqei-bootstrap-identity-admin = "yuqei_contract_api.identity_bootstrap:main"
 yuqei-public-policy-promotion = "yuqei_contract_api.public_policy_promotion_cli:main"
 

yuqei-contract-platform/services/contract-api/src/yuqei_contract_api/v1_legal_fulltext_import.py

@@ -1,798 +0,0 @@
-from __future__ import annotations
-
-import argparse
-import json
-from collections.abc import Iterable, Iterator
-from dataclasses import dataclass, field
-from datetime import UTC, datetime
-from hashlib import sha256
-from pathlib import Path
-from typing import Any
-
-import psycopg
-from psycopg import sql
-
-from yuqei_contract_api.config import ContractApiSettings
-
-
-DEFAULT_VERSION = "v1-legal-20260616"
-DEFAULT_ACTOR = "v1-legal-fulltext-import"
-DEFAULT_TENANT_ID = "default"
-
-TARGET_COLUMNS = [
-    "id",
-    "tenant_id",
-    "article_id",
-    "version",
-    "source_id",
-    "source_aliases",
-    "source_title",
-    "source_type",
-    "authority",
-    "article_no",
-    "title",
-    "chapter_title",
-    "section_title",
-    "article_text",
-    "summary",
-    "keywords",
-    "tags",
-    "status",
-    "effective_date",
-    "source_url",
-    "published_at",
-    "created_by",
-    "updated_by",
-    "deleted_at",
-]
-
-SOURCE_DATABASE_COLUMNS = [
-    "id",
-    "tenant_id",
-    "source_id",
-    "source_title",
-    "source_type",
-    "authority",
-    "effective_date",
-    "validity_status",
-    "official_url",
-    "article_no",
-    "chapter_title",
-    "section_title",
-    "article_text",
-    # Some legacy V1 rows contain non-UTF-8 JSON bytes in tags_json. Tags are
-    # supplemental only, so avoid decoding them during the fulltext import.
-    "NULL::text AS tags_json",
-    "published_at",
-    "created_at",
-    "updated_at",
-]
-
-SOURCE_DATABASE_COLUMN_SQL = sql.SQL(", ").join(sql.SQL(column) for column in SOURCE_DATABASE_COLUMNS)
-LEGAL_ARTICLES_TABLE_SQL = sql.Identifier("legal_articles")
-SOURCE_ROWS_TABLE_SQL = sql.Identifier("source_rows")
-
-
-@dataclass(frozen=True)
-class V1LegalArticle:
-    source_article_id: str | None
-    tenant_id: str
-    source_id: str
-    source_title: str
-    source_type: str
-    authority: str | None
-    article_no: str
-    chapter_title: str | None = None
-    section_title: str | None = None
-    article_text: str = ""
-    tags: list[str] = field(default_factory=list)
-    validity_status: str | None = None
-    effective_date: str | None = None
-    source_url: str | None = None
-    published_at: datetime | None = None
-
-
-@dataclass(frozen=True)
-class ImportSummary:
-    command: str
-    mode: str
-    source_kind: str
-    version: str
-    source_count: int
-    with_text_count: int
-    metadata_only_count: int
-    source_offset: int = 0
-    inserted_count: int = 0
-    updated_count: int = 0
-    batches: int = 0
-    target_before_count: int | None = None
-    target_after_count: int | None = None
-    sample_article_ids: list[str] = field(default_factory=list)
-
-    def as_payload(self) -> dict[str, Any]:
-        return {
-            "command": self.command,
-            "mode": self.mode,
-            "source_kind": self.source_kind,
-            "version": self.version,
-            "source_offset": self.source_offset,
-            "source_count": self.source_count,
-            "with_text_count": self.with_text_count,
-            "metadata_only_count": self.metadata_only_count,
-            "inserted_count": self.inserted_count,
-            "updated_count": self.updated_count,
-            "batches": self.batches,
-            "target_before_count": self.target_before_count,
-            "target_after_count": self.target_after_count,
-            "sample_article_ids": self.sample_article_ids,
-        }
-
-
-def import_from_v1_database(
-    *,
-    source_database_url: str,
-    target_database_url: str,
-    version: str = DEFAULT_VERSION,
-    apply: bool = False,
-    batch_size: int = 5000,
-    offset: int = 0,
-    after_source_id: str | None = None,
-    after_article_no: str | None = None,
-    after_source_article_id: str | None = None,
-    limit: int | None = None,
-    actor_name: str = DEFAULT_ACTOR,
-) -> ImportSummary:
-    offset = max(0, offset)
-    after_key = _after_key(after_source_id, after_article_no, after_source_article_id)
-    source_dsn = normalize_database_url(source_database_url)
-    target_dsn = normalize_database_url(target_database_url)
-    with psycopg.connect(source_dsn) as source_conn, psycopg.connect(target_dsn) as target_conn:
-        source_count, with_text_count = _source_database_counts(
-            source_conn,
-            offset=offset,
-            limit=limit,
-            after_key=after_key,
-        )
-        target_before = _target_version_count(target_conn, version=version)
-        if not apply:
-            samples = _source_database_samples(
-                source_conn,
-                version=version,
-                offset=offset,
-                limit=min(5, limit or 5),
-                after_key=after_key,
-            )
-            return ImportSummary(
-                command="import-v1-legal-fulltext",
-                mode="dry-run",
-                source_kind="v1-database",
-                version=version,
-                source_offset=offset,
-                source_count=source_count,
-                with_text_count=with_text_count,
-                metadata_only_count=source_count - with_text_count,
-                target_before_count=target_before,
-                sample_article_ids=samples,
-            )
-
-        _create_staging_table(target_conn)
-        inserted_count = 0
-        updated_count = 0
-        batches = 0
-        for batch in _iter_source_database_articles(
-            source_conn,
-            version=version,
-            batch_size=batch_size,
-            offset=offset,
-            after_key=after_key,
-            limit=limit,
-        ):
-            result = _import_batch(target_conn, batch, actor_name=actor_name)
-            inserted_count += result["inserted"]
-            updated_count += result["updated"]
-            batches += 1
-        target_after = _target_version_count(target_conn, version=version)
-        return ImportSummary(
-            command="import-v1-legal-fulltext",
-            mode="apply",
-            source_kind="v1-database",
-            version=version,
-            source_offset=offset,
-            source_count=source_count,
-            with_text_count=with_text_count,
-            metadata_only_count=source_count - with_text_count,
-            inserted_count=inserted_count,
-            updated_count=updated_count,
-            batches=batches,
-            target_before_count=target_before,
-            target_after_count=target_after,
-        )
-
-
-def import_from_phase14_directory(
-    *,
-    source_dir: Path,
-    target_database_url: str,
-    version: str = DEFAULT_VERSION,
-    apply: bool = False,
-    batch_size: int = 5000,
-    offset: int = 0,
-    limit: int | None = None,
-    actor_name: str = DEFAULT_ACTOR,
-) -> ImportSummary:
-    target_dsn = normalize_database_url(target_database_url)
-    articles = list(iter_phase14_articles(source_dir, version=version, offset=offset, limit=limit))
-    source_count = len(articles)
-    with_text_count = sum(1 for item in articles if item["article_text"])
-    with psycopg.connect(target_dsn) as target_conn:
-        target_before = _target_version_count(target_conn, version=version)
-        if not apply:
-            return ImportSummary(
-                command="import-v1-legal-fulltext",
-                mode="dry-run",
-                source_kind="phase14-directory",
-                version=version,
-                source_offset=max(0, offset),
-                source_count=source_count,
-                with_text_count=with_text_count,
-                metadata_only_count=source_count - with_text_count,
-                target_before_count=target_before,
-                sample_article_ids=[str(row["article_id"]) for row in articles[:5]],
-            )
-        _create_staging_table(target_conn)
-        inserted_count = 0
-        updated_count = 0
-        batches = 0
-        for batch in batched(articles, batch_size):
-            result = _import_batch(target_conn, batch, actor_name=actor_name)
-            inserted_count += result["inserted"]
-            updated_count += result["updated"]
-            batches += 1
-        target_after = _target_version_count(target_conn, version=version)
-    return ImportSummary(
-        command="import-v1-legal-fulltext",
-        mode="apply",
-        source_kind="phase14-directory",
-        version=version,
-        source_offset=max(0, offset),
-        source_count=source_count,
-        with_text_count=with_text_count,
-        metadata_only_count=source_count - with_text_count,
-        inserted_count=inserted_count,
-        updated_count=updated_count,
-        batches=batches,
-        target_before_count=target_before,
-        target_after_count=target_after,
-    )
-
-
-def iter_phase14_articles(
-    source_dir: Path,
-    *,
-    version: str,
-    offset: int = 0,
-    limit: int | None = None,
-) -> Iterator[dict[str, Any]]:
-    seen: set[tuple[str, str, str, str]] = set()
-    skipped = 0
-    emitted = 0
-    offset = max(0, offset)
-    for path in _phase14_manifest_files(source_dir):
-        payload = json.loads(path.read_text(encoding="utf-8-sig"))
-        for source in _iter_sources(payload):
-            source_id = _text(source.get("source_id")) or _hash_id("source", _text(source.get("title")) or path.stem)
-            source_title = _text(source.get("title")) or source_id
-            source_type = _text(source.get("source_type")) or "law"
-            authority = _text(source.get("authority"))
-            effective_date = _short_date(source.get("effective_date"))
-            source_url = _text(source.get("source_url")) or _text(source.get("official_url")) or _text(source.get("homepage"))
-            for article in source.get("articles") or []:
-                if not isinstance(article, dict):
-                    continue
-                article_no = _text(article.get("article_no"))
-                if not article_no:
-                    continue
-                chapter_title = _text(article.get("chapter_title"))
-                section_title = _text(article.get("section_title"))
-                dedupe_key = (source_id, article_no, chapter_title or "", section_title or "")
-                if dedupe_key in seen:
-                    continue
-                seen.add(dedupe_key)
-                tags = _string_list(article.get("tags"))
-                article_text = _text(
-                    article.get("article_text")
-                    or article.get("text")
-                    or article.get("content")
-                    or article.get("full_text")
-                ) or ""
-                if not article_text:
-                    tags = [*tags, "metadata-only"]
-                item = V1LegalArticle(
-                    source_article_id=None,
-                    tenant_id=DEFAULT_TENANT_ID,
-                    source_id=source_id,
-                    source_title=source_title,
-                    source_type=source_type,
-                    authority=authority,
-                    article_no=article_no,
-                    chapter_title=chapter_title,
-                    section_title=section_title,
-                    article_text=article_text,
-                    tags=[*tags, "v1-phase14"],
-                    validity_status=_text(source.get("validity_status")),
-                    effective_date=effective_date,
-                    source_url=source_url,
-                )
-                if skipped < offset:
-                    skipped += 1
-                    continue
-                yield _target_row(item, version=version)
-                emitted += 1
-                if limit is not None and emitted >= limit:
-                    return
-
-
-def normalize_database_url(database_url: str) -> str:
-    return database_url.replace("postgresql+psycopg://", "postgresql://", 1)
-
-
-def batched(items: Iterable[dict[str, Any]], batch_size: int) -> Iterator[list[dict[str, Any]]]:
-    batch: list[dict[str, Any]] = []
-    for item in items:
-        batch.append(item)
-        if len(batch) >= batch_size:
-            yield batch
-            batch = []
-    if batch:
-        yield batch
-
-
-def build_parser() -> argparse.ArgumentParser:
-    parser = argparse.ArgumentParser(description="Import V1 legal articles into V2 legal_fulltext_versions.")
-    parser.add_argument("--database-url", default=None, help="Target V2 CONTRACT_API_DATABASE_URL.")
-    parser.add_argument("--source-database-url", default=None, help="V1 database URL containing legal_articles.")
-    parser.add_argument("--source-dir", type=Path, default=None, help="V1 Phase14 legal-fulltext-library directory.")
-    parser.add_argument("--version", default=DEFAULT_VERSION)
-    parser.add_argument("--batch-size", type=int, default=5000)
-    parser.add_argument("--offset", type=int, default=0, help="Skip this many ordered source rows before importing.")
-    parser.add_argument("--after-source-id", default=None, help="Resume after this V1 source_id in source order.")
-    parser.add_argument("--after-article-no", default=None, help="Resume after this V1 article_no in source order.")
-    parser.add_argument("--after-source-article-id", default=None, help="Resume after this V1 legal_articles.id.")
-    parser.add_argument("--limit", type=int, default=None)
-    parser.add_argument("--apply", action="store_true", help="Write target rows. Omit for dry-run.")
-    parser.add_argument("--actor-name", default=DEFAULT_ACTOR)
-    return parser
-
-
-def main() -> None:
-    args = build_parser().parse_args()
-    target_database_url = args.database_url or ContractApiSettings().database_url
-    if args.source_database_url and args.source_dir:
-        raise SystemExit("--source-database-url and --source-dir are mutually exclusive.")
-    if args.source_database_url:
-        result = import_from_v1_database(
-            source_database_url=args.source_database_url,
-            target_database_url=target_database_url,
-            version=args.version,
-            apply=args.apply,
-            batch_size=args.batch_size,
-            offset=args.offset,
-            after_source_id=args.after_source_id,
-            after_article_no=args.after_article_no,
-            after_source_article_id=args.after_source_article_id,
-            limit=args.limit,
-            actor_name=args.actor_name,
-        )
-    elif args.source_dir:
-        result = import_from_phase14_directory(
-            source_dir=args.source_dir,
-            target_database_url=target_database_url,
-            version=args.version,
-            apply=args.apply,
-            batch_size=args.batch_size,
-            offset=args.offset,
-            limit=args.limit,
-            actor_name=args.actor_name,
-        )
-    else:
-        raise SystemExit("One of --source-database-url or --source-dir is required.")
-    print(json.dumps(result.as_payload(), ensure_ascii=False, indent=2))
-
-
-def _source_database_counts(
-    source_conn: psycopg.Connection,
-    *,
-    offset: int = 0,
-    limit: int | None,
-    after_key: tuple[str, str, str] | None = None,
-) -> tuple[int, int]:
-    where_sql = sql.SQL("WHERE (source_id, article_no, id) > (%s, %s, %s)") if after_key else sql.SQL("")
-    limit_sql = sql.SQL("LIMIT {limit}").format(limit=sql.Literal(int(limit))) if limit is not None else sql.SQL("")
-    offset_sql = sql.SQL("OFFSET {offset}").format(offset=sql.Literal(int(max(0, offset)))) if offset else sql.SQL("")
-    query = sql.SQL(
-        """
-        WITH {source_rows} AS (
-            SELECT article_text
-            FROM {legal_articles}
-            {where_clause}
-            ORDER BY source_id, article_no, id
-            {limit_clause}
-            {offset_clause}
-        )
-        SELECT count(*), count(*) FILTER (WHERE length(coalesce(article_text, '')) > 0)
-        FROM {source_rows}
-        """
-    ).format(
-        source_rows=SOURCE_ROWS_TABLE_SQL,
-        legal_articles=LEGAL_ARTICLES_TABLE_SQL,
-        where_clause=where_sql,
-        limit_clause=limit_sql,
-        offset_clause=offset_sql,
-    )
-    with source_conn.cursor() as cursor:
-        cursor.execute(query, after_key or ())
-        row = cursor.fetchone()
-    return int(row[0] or 0), int(row[1] or 0)
-
-
-def _source_database_samples(
-    source_conn: psycopg.Connection,
-    *,
-    version: str,
-    offset: int = 0,
-    limit: int = 5,
-    after_key: tuple[str, str, str] | None = None,
-) -> list[str]:
-    where_sql = sql.SQL("WHERE (source_id, article_no, id) > (%s, %s, %s)") if after_key else sql.SQL("")
-    params: list[Any] = [*(after_key or ()), max(0, offset), limit]
-    with source_conn.cursor(row_factory=psycopg.rows.dict_row) as cursor:
-        cursor.execute(
-            sql.SQL(
-                """
-                SELECT {columns}
-                FROM {legal_articles}
-                {where_clause}
-                ORDER BY source_id, article_no, id
-                OFFSET %s
-                LIMIT %s
-                """
-            ).format(
-                columns=SOURCE_DATABASE_COLUMN_SQL,
-                legal_articles=LEGAL_ARTICLES_TABLE_SQL,
-                where_clause=where_sql,
-            ),
-            params,
-        )
-        return [str(_target_row(_article_from_source_database_row(row), version=version)["article_id"]) for row in cursor]
-
-
-def _iter_source_database_articles(
-    source_conn: psycopg.Connection,
-    *,
-    version: str,
-    batch_size: int,
-    offset: int,
-    after_key: tuple[str, str, str] | None,
-    limit: int | None,
-) -> Iterator[list[dict[str, Any]]]:
-    where_sql = sql.SQL("WHERE (source_id, article_no, id) > (%s, %s, %s)") if after_key else sql.SQL("")
-    query = sql.SQL(
-        """
-        SELECT {columns}
-        FROM {legal_articles}
-        {where_clause}
-        ORDER BY source_id, article_no, id
-        """
-    ).format(
-        columns=SOURCE_DATABASE_COLUMN_SQL,
-        legal_articles=LEGAL_ARTICLES_TABLE_SQL,
-        where_clause=where_sql,
-    )
-    if limit is not None:
-        query = sql.SQL("{} LIMIT {}").format(query, sql.Literal(int(limit)))
-    if offset:
-        query = sql.SQL("{} OFFSET {}").format(query, sql.Literal(int(max(0, offset))))
-    with source_conn.cursor(name="v1_legal_articles_import", row_factory=psycopg.rows.dict_row) as cursor:
-        cursor.itersize = batch_size
-        cursor.execute(query, after_key or ())
-        while True:
-            rows = cursor.fetchmany(batch_size)
-            if not rows:
-                break
-            yield [_target_row(_article_from_source_database_row(row), version=version) for row in rows]
-
-
-def _after_key(
-    source_id: str | None,
-    article_no: str | None,
-    source_article_id: str | None,
-) -> tuple[str, str, str] | None:
-    values = (_text(source_id), _text(article_no), _text(source_article_id))
-    if not any(values):
-        return None
-    if not all(values):
-        raise ValueError("--after-source-id, --after-article-no and --after-source-article-id must be provided together.")
-    return values  # type: ignore[return-value]
-
-
-def _article_from_source_database_row(row: dict[str, Any]) -> V1LegalArticle:
-    return V1LegalArticle(
-        source_article_id=_text(row.get("id")),
-        tenant_id=DEFAULT_TENANT_ID,
-        source_id=_text(row.get("source_id")) or _text(row.get("source_title")) or "unknown-source",
-        source_title=_text(row.get("source_title")) or _text(row.get("source_id")) or "Unknown Legal Source",
-        source_type=_text(row.get("source_type")) or "law",
-        authority=_text(row.get("authority")),
-        article_no=_text(row.get("article_no")) or "unknown-article",
-        chapter_title=_text(row.get("chapter_title")),
-        section_title=_text(row.get("section_title")),
-        article_text=_text(row.get("article_text")) or "",
-        tags=_string_list(row.get("tags_json")),
-        validity_status=_text(row.get("validity_status")),
-        effective_date=_short_date(row.get("effective_date")),
-        source_url=_text(row.get("official_url")),
-        published_at=_datetime_or_none(row.get("published_at")) or _datetime_or_none(row.get("updated_at")),
-    )
-
-
-def _target_row(article: V1LegalArticle, *, version: str) -> dict[str, Any]:
-    source_aliases = [article.source_id]
-    if article.source_article_id:
-        source_aliases.append(article.source_article_id)
-    source_id = _canonical_id(article.source_id, prefix="v1-source", max_length=128)
-    article_id = _hash_id("v1-law", article.tenant_id, article.source_id, article.article_no, article.source_article_id or "")
-    tags = _unique_strings([*article.tags, "v1-legal-library"])
-    if not article.article_text:
-        tags = _unique_strings([*tags, "metadata-only"])
-    published_at = article.published_at or datetime.now(UTC)
-    return {
-        "id": _hash_id("lfv", article.tenant_id, article_id, version),
-        "tenant_id": _clip(article.tenant_id or DEFAULT_TENANT_ID, 64),
-        "article_id": _clip(article_id, 128),
-        "version": _clip(version, 128),
-        "source_id": _clip(source_id, 128),
-        "source_aliases": source_aliases,
-        "source_title": _clip(article.source_title, 255),
-        "source_type": _clip(article.source_type or "law", 64),
-        "authority": _clip(article.authority, 255),
-        "article_no": _clip(article.article_no, 128),
-        "title": _clip(f"{article.source_title} {article.article_no}", 255),
-        "chapter_title": _clip(article.chapter_title, 255),
-        "section_title": _clip(article.section_title, 255),
-        "article_text": article.article_text,
-        "summary": _summary_for(article),
-        "keywords": [],
-        "tags": tags,
-        "status": _map_status(article.validity_status),
-        "effective_date": _short_date(article.effective_date),
-        "source_url": article.source_url,
-        "published_at": published_at,
-        "created_by": DEFAULT_ACTOR,
-        "updated_by": DEFAULT_ACTOR,
-        "deleted_at": None,
-    }
-
-
-def _create_staging_table(conn: psycopg.Connection) -> None:
-    with conn.cursor() as cursor:
-        cursor.execute(
-            """
-            CREATE TEMP TABLE IF NOT EXISTS legal_fulltext_versions_import_stage (
-                id text,
-                tenant_id text,
-                article_id text,
-                version text,
-                source_id text,
-                source_aliases json,
-                source_title text,
-                source_type text,
-                authority text,
-                article_no text,
-                title text,
-                chapter_title text,
-                section_title text,
-                article_text text,
-                summary text,
-                keywords json,
-                tags json,
-                status text,
-                effective_date text,
-                source_url text,
-                published_at timestamptz,
-                created_by text,
-                updated_by text,
-                deleted_at timestamptz
-            ) ON COMMIT PRESERVE ROWS
-            """
-        )
-        conn.commit()
-
-
-def _import_batch(conn: psycopg.Connection, batch: list[dict[str, Any]], *, actor_name: str) -> dict[str, int]:
-    if not batch:
-        return {"inserted": 0, "updated": 0}
-    with conn.cursor() as cursor:
-        cursor.execute("TRUNCATE legal_fulltext_versions_import_stage")
-        copy_sql = f"COPY legal_fulltext_versions_import_stage ({', '.join(TARGET_COLUMNS)}) FROM STDIN"
-        with cursor.copy(copy_sql) as copy:
-            for row in batch:
-                copy.write_row(_copy_row(row, actor_name=actor_name))
-        cursor.execute(
-            """
-            WITH upserted AS (
-                INSERT INTO legal_fulltext_versions (
-                    id, tenant_id, article_id, version, source_id, source_aliases,
-                    source_title, source_type, authority, article_no, title,
-                    chapter_title, section_title, article_text, summary, keywords,
-                    tags, status, effective_date, source_url, published_at,
-                    created_by, updated_by, deleted_at
-                )
-                SELECT
-                    id, tenant_id, article_id, version, source_id, source_aliases,
-                    source_title, source_type, nullif(authority, ''), article_no, title,
-                    nullif(chapter_title, ''), nullif(section_title, ''), article_text,
-                    summary, keywords, tags, status, nullif(effective_date, ''),
-                    nullif(source_url, ''), published_at, created_by, updated_by, deleted_at
-                FROM legal_fulltext_versions_import_stage
-                ON CONFLICT (tenant_id, article_id, version) DO UPDATE SET
-                    source_id = EXCLUDED.source_id,
-                    source_aliases = EXCLUDED.source_aliases,
-                    source_title = EXCLUDED.source_title,
-                    source_type = EXCLUDED.source_type,
-                    authority = EXCLUDED.authority,
-                    article_no = EXCLUDED.article_no,
-                    title = EXCLUDED.title,
-                    chapter_title = EXCLUDED.chapter_title,
-                    section_title = EXCLUDED.section_title,
-                    article_text = EXCLUDED.article_text,
-                    summary = EXCLUDED.summary,
-                    keywords = EXCLUDED.keywords,
-                    tags = EXCLUDED.tags,
-                    status = EXCLUDED.status,
-                    effective_date = EXCLUDED.effective_date,
-                    source_url = EXCLUDED.source_url,
-                    published_at = EXCLUDED.published_at,
-                    updated_by = EXCLUDED.updated_by,
-                    deleted_at = EXCLUDED.deleted_at
-                RETURNING (xmax = 0) AS inserted
-            )
-            SELECT
-                count(*) FILTER (WHERE inserted),
-                count(*) FILTER (WHERE NOT inserted)
-            FROM upserted
-            """
-        )
-        inserted, updated = cursor.fetchone()
-        conn.commit()
-    return {"inserted": int(inserted or 0), "updated": int(updated or 0)}
-
-
-def _copy_row(row: dict[str, Any], *, actor_name: str) -> list[Any]:
-    values: list[Any] = []
-    for column in TARGET_COLUMNS:
-        value = row.get(column)
-        if column in {"source_aliases", "keywords", "tags"}:
-            value = json.dumps(value or [], ensure_ascii=False)
-        elif column in {"created_by", "updated_by"}:
-            value = actor_name
-        values.append(value)
-    return values
-
-
-def _target_version_count(conn: psycopg.Connection, *, version: str) -> int:
-    with conn.cursor() as cursor:
-        cursor.execute("SELECT count(*) FROM legal_fulltext_versions WHERE version = %s", (version,))
-        return int(cursor.fetchone()[0] or 0)
-
-
-def _phase14_manifest_files(source_dir: Path) -> list[Path]:
-    candidates = [
-        source_dir / "legal-article-manifest.json",
-        source_dir / "legal-article-expansion-batch-001.json",
-        source_dir / "priority1-law-import-manifest.json",
-        *sorted((source_dir / "incremental-batches").glob("*.json")),
-    ]
-    return [path for path in candidates if path.exists() and path.stat().st_size > 0]
-
-
-def _iter_sources(payload: Any) -> Iterator[dict[str, Any]]:
-    if isinstance(payload, dict):
-        sources = payload.get("sources")
-        if isinstance(sources, list):
-            for source in sources:
-                if isinstance(source, dict):
-                    yield source
-        for value in payload.values():
-            if isinstance(value, (dict, list)):
-                yield from _iter_sources(value)
-    elif isinstance(payload, list):
-        for value in payload:
-            yield from _iter_sources(value)
-
-
-def _summary_for(article: V1LegalArticle) -> str:
-    if article.article_text:
-        excerpt = article.article_text.replace("\n", " ").strip()[:180]
-        return f"V1法律全文库迁移：{article.source_title} {article.article_no}。{excerpt}"
-    return (
-        "V1 Phase14 metadata-only import: source manifest included article metadata "
-        "but no article_text/full_text payload."
-    )
-
-
-def _map_status(validity_status: str | None) -> str:
-    normalized = (validity_status or "").strip().lower()
-    if any(token in normalized for token in ("废止", "失效", "repealed", "invalid")):
-        return "repealed"
-    if any(token in normalized for token in ("修订", "修改", "amended")):
-        return "amended"
-    return "effective"
-
-
-def _canonical_id(value: str, *, prefix: str, max_length: int) -> str:
-    text = _text(value) or prefix
-    if len(text) <= max_length:
-        return text
-    return _hash_id(prefix, text)
-
-
-def _hash_id(prefix: str, *parts: object) -> str:
-    digest = sha256("|".join(str(part or "") for part in parts).encode("utf-8")).hexdigest()
-    return f"{prefix}-{digest[:32]}"
-
-
-def _clip(value: object, max_length: int) -> str | None:
-    text = _text(value)
-    if text is None:
-        return None
-    return text[:max_length]
-
-
-def _short_date(value: object) -> str | None:
-    text = _text(value)
-    if not text:
-        return None
-    return text[:32]
-
-
-def _text(value: object) -> str | None:
-    if value is None:
-        return None
-    if isinstance(value, datetime):
-        return value.date().isoformat()
-    text = str(value).strip()
-    return text or None
-
-
-def _datetime_or_none(value: object) -> datetime | None:
-    if isinstance(value, datetime):
-        return value if value.tzinfo else value.replace(tzinfo=UTC)
-    return None
-
-
-def _string_list(value: object) -> list[str]:
-    if value is None:
-        return []
-    if isinstance(value, str):
-        try:
-            parsed = json.loads(value)
-        except json.JSONDecodeError:
-            parsed = [value]
-        return _string_list(parsed)
-    if isinstance(value, list):
-        return _unique_strings(str(item).strip() for item in value if str(item).strip())
-    return []
-
-
-def _unique_strings(values: Iterable[str]) -> list[str]:
-    seen: set[str] = set()
-    result: list[str] = []
-    for value in values:
-        if value not in seen:
-            seen.add(value)
-            result.append(value)
-    return result
-
-
-if __name__ == "__main__":
-    main()

yuqei-contract-platform/services/contract-api/tests/test_v1_legal_fulltext_import.py

@@ -1,127 +0,0 @@
-import json
-from datetime import UTC, datetime
-from pathlib import Path
-
-from yuqei_contract_api.v1_legal_fulltext_import import (
-    _article_from_source_database_row,
-    _map_status,
-    _target_row,
-    iter_phase14_articles,
-)
-
-
-def test_phase14_directory_rows_include_metadata_only_tag(tmp_path: Path) -> None:
-    source_dir = tmp_path / "phase14-legal-fulltext-library"
-    source_dir.mkdir()
-    (source_dir / "legal-article-manifest.json").write_text(
-        json.dumps(
-            {
-                "sources": [
-                    {
-                        "source_id": "civil-code",
-                        "title": "Civil Code",
-                        "source_type": "law",
-                        "authority": "NPC",
-                        "validity_status": "effective",
-                        "articles": [
-                            {
-                                "article_no": "Article 1",
-                                "article_text": "Civil law text.",
-                                "tags": ["contract"],
-                            },
-                            {
-                                "article_no": "Article 2",
-                                "chapter_title": "General",
-                            },
-                        ],
-                    }
-                ]
-            }
-        ),
-        encoding="utf-8",
-    )
-
-    rows = list(iter_phase14_articles(source_dir, version="test-version"))
-
-    assert len(rows) == 2
-    assert rows[0]["version"] == "test-version"
-    assert rows[0]["source_id"] == "civil-code"
-    assert rows[0]["article_text"] == "Civil law text."
-    assert "Civil Code" in rows[0]["summary"]
-    assert "metadata-only" not in rows[0]["tags"]
-    assert rows[1]["article_text"] == ""
-    assert "metadata-only" in rows[1]["tags"]
-    assert "v1-phase14" in rows[1]["tags"]
-    assert "v1-legal-library" in rows[1]["tags"]
-
-
-def test_phase14_directory_supports_offset_and_limit(tmp_path: Path) -> None:
-    source_dir = tmp_path / "phase14-legal-fulltext-library"
-    source_dir.mkdir()
-    (source_dir / "legal-article-manifest.json").write_text(
-        json.dumps(
-            {
-                "sources": [
-                    {
-                        "source_id": "civil-code",
-                        "title": "Civil Code",
-                        "articles": [
-                            {"article_no": "Article 1", "article_text": "one"},
-                            {"article_no": "Article 2", "article_text": "two"},
-                            {"article_no": "Article 3", "article_text": "three"},
-                        ],
-                    }
-                ]
-            }
-        ),
-        encoding="utf-8",
-    )
-
-    rows = list(iter_phase14_articles(source_dir, version="test-version", offset=1, limit=1))
-
-    assert len(rows) == 1
-    assert rows[0]["article_no"] == "Article 2"
-    assert rows[0]["article_text"] == "two"
-
-
-def test_source_database_row_maps_status_and_stable_target_fields() -> None:
-    row = {
-        "id": "v1-row-1",
-        "tenant_id": None,
-        "source_id": "law-source",
-        "source_title": "Example Law",
-        "source_type": None,
-        "authority": "Authority",
-        "effective_date": "2026-01-01",
-        "validity_status": "\u5df2\u5e9f\u6b62",
-        "official_url": "https://example.test/law",
-        "article_no": "Article 9",
-        "chapter_title": "Chapter",
-        "section_title": "Section",
-        "article_text": "Legal body",
-        "tags_json": "[\"sample\"]",
-        "published_at": datetime(2026, 1, 2, tzinfo=UTC),
-        "created_at": None,
-        "updated_at": None,
-    }
-
-    article = _article_from_source_database_row(row)
-    target = _target_row(article, version="v1-test")
-
-    assert target["tenant_id"] == "default"
-    assert target["version"] == "v1-test"
-    assert target["source_id"] == "law-source"
-    assert target["source_aliases"] == ["law-source", "v1-row-1"]
-    assert target["source_type"] == "law"
-    assert target["status"] == "repealed"
-    assert target["tags"] == ["sample", "v1-legal-library"]
-    assert "Example Law" in target["summary"]
-    assert target["article_id"].startswith("v1-law-")
-
-
-def test_map_status_handles_chinese_and_english_statuses() -> None:
-    assert _map_status("\u5df2\u5e9f\u6b62") == "repealed"
-    assert _map_status("\u5df2\u5931\u6548") == "repealed"
-    assert _map_status("2024\u5e74\u4fee\u8ba2") == "amended"
-    assert _map_status("amended") == "amended"
-    assert _map_status("effective") == "effective"

yuqei-infra/README.md

@@ -61,7 +61,7 @@ The script checks:
 
 ## Remote Smoke Deploy
 
-Use this before replacing the old `contact.yuqei.com` deployment. It deploys V2 to an independent directory and independent ports on `192.168.3.90`, so it does not overwrite the V1 backend or the Aliyun frontend.
+Use this to validate the V2 stack in an isolated directory and on independent ports at `192.168.3.90` before any separately authorized production promotion.
 
 Default remote path:
 
@@ -105,7 +105,7 @@ Latest passing smoke artifact:
 - Remote database baseline passed for `yuqei_contract`, `yuqei_ai`, and `yuqei_ops`.
 - Contract API runs with one Uvicorn worker and a 2-second read cache; all mutating requests invalidate the cache before and after execution.
 
-After smoke passes, the old production deployment can be replaced deliberately: backend/AI/worker/database/MinIO on `192.168.3.90`, frontend on Aliyun, and `contact.yuqei.com` kept as the public entry.
+After smoke passes, production promotion remains a deliberate, separately authorized action: backend/AI/worker/database/MinIO on `192.168.3.90`, frontend on Aliyun, and `contact.yuqei.com` kept as the public entry.
 
 Production is now a separate, guarded path. Use `compose/docker-compose.production.yml` together with the shared compose file, then follow `deploy/README.md`. The production path requires digest-pinned images, loopback-only internal ports, TLS edge configuration, an off-host backup mount, protected provider secret files, and a passing `deploy/preflight-production.py`. It intentionally has no automatic traffic-cutover command.
 

yuqei-infra/compose/docker-compose.yml

@@ -244,7 +244,7 @@ services:
       args:
         PIP_INDEX_URL: ${PYPI_INDEX_URL:-https://pypi.org/simple}
     working_dir: /app/yuqei-ai-platform/services/ai-platform-api
-    command: ["python", "-m", "yuqei_ai_platform_api.v1_prompt_assets_import", "--apply"]
+    command: ["python", "-m", "yuqei_ai_platform_api.packaged_prompt_assets_import", "--apply"]
     environment:
       AI_PLATFORM_ENVIRONMENT: ${AI_PLATFORM_ENVIRONMENT:-local}
       AI_PLATFORM_DATABASE_URL: ${AI_PLATFORM_DATABASE_URL:-postgresql+psycopg://yuqei:yuqei-local@postgres:5432/yuqei_ai}

yuqei-infra/scripts/compare-v1-v2-assets.py

@@ -1,347 +0,0 @@
-"""Compare V1 source assets with V2 packaged and runtime assets."""
-
-from __future__ import annotations
-
-import argparse
-import json
-import os
-from datetime import UTC, datetime
-from pathlib import Path
-from typing import Any
-from urllib.error import URLError
-from urllib.request import Request, urlopen
-
-
-REPO_ROOT = Path(__file__).resolve().parents[2]
-DEFAULT_V1_ROOT = Path("D:/yuqei-project/ai-contract-platform-v1")
-DEFAULT_CONTRACT_API_BASE = "http://localhost:8101"
-DEFAULT_AI_PLATFORM_API_BASE = "http://localhost:8102"
-
-
-def main() -> int:
-    parser = argparse.ArgumentParser(description="Compare V1 and V2 release assets.")
-    parser.add_argument("--v1-root", type=Path, default=DEFAULT_V1_ROOT)
-    parser.add_argument("--contract-api-base", default=DEFAULT_CONTRACT_API_BASE)
-    parser.add_argument("--ai-platform-api-base", default=DEFAULT_AI_PLATFORM_API_BASE)
-    parser.add_argument("--tenant-id", default="default")
-    parser.add_argument("--user-id", default="user-current-admin")
-    parser.add_argument("--contract-proxy-key-env", default="CONTRACT_API_WEB_PROXY_SHARED_SECRET")
-    parser.add_argument("--ai-service-key-env", default="AI_PLATFORM_SERVICE_SHARED_SECRET")
-    parser.add_argument(
-        "--artifact-dir",
-        type=Path,
-        default=REPO_ROOT / "project-docs" / "qa-artifacts" / "v2-asset-count",
-    )
-    args = parser.parse_args()
-
-    checked_at = datetime.now(UTC).isoformat()
-    artifact_dir = args.artifact_dir
-    artifact_dir.mkdir(parents=True, exist_ok=True)
-    stamp = checked_at.replace(":", "-").replace("+00:00", "Z")
-    contract_api_base = args.contract_api_base.rstrip("/")
-    ai_platform_api_base = args.ai_platform_api_base.rstrip("/")
-    contract_headers = identity_headers(args.tenant_id, args.user_id)
-    ai_platform_headers = identity_headers(args.tenant_id, args.user_id)
-    contract_proxy_key = os.getenv(args.contract_proxy_key_env, "").strip()
-    ai_service_key = os.getenv(args.ai_service_key_env, "").strip()
-    if contract_proxy_key:
-        contract_headers["X-Contract-Proxy-Key"] = contract_proxy_key
-    if ai_service_key:
-        ai_platform_headers["X-AI-Platform-Service-Key"] = ai_service_key
-
-    counts, parity = build_asset_report(
-        args.v1_root,
-        contract_api_base,
-        ai_platform_api_base,
-        contract_headers=contract_headers,
-        ai_platform_headers=ai_platform_headers,
-    )
-    result = {
-        "status": derive_status(counts, parity),
-        "checked_at": checked_at,
-        "v1_root": str(args.v1_root),
-        "contract_api_base": contract_api_base,
-        "ai_platform_api_base": ai_platform_api_base,
-        "request_context": {
-            "tenant_id": args.tenant_id,
-            "user_id": args.user_id,
-            "contract_proxy_auth_configured": bool(contract_proxy_key),
-            "ai_service_auth_configured": bool(ai_service_key),
-        },
-        "counts": counts,
-        "parity": parity,
-        "notes": [
-            "Template, clause, rule, and Prompt parity compares stable IDs, not only file counts.",
-            "Runtime totals depend on the target databases behind the supplied API bases.",
-            "Semantic legal review and model-output quality still require sampled human approval.",
-        ],
-    }
-
-    json_path = artifact_dir / f"v2-asset-count-{stamp}.json"
-    md_path = artifact_dir / f"v2-asset-count-{stamp}.md"
-    json_path.write_text(json.dumps(result, ensure_ascii=False, indent=2), encoding="utf-8")
-    md_path.write_text(render_markdown(result), encoding="utf-8")
-    print(
-        json.dumps(
-            {"status": result["status"], "json": str(json_path), "markdown": str(md_path)},
-            ensure_ascii=False,
-        )
-    )
-    return 0 if result["status"] in {"pass", "warning"} else 1
-
-
-def build_asset_report(
-    v1_root: Path,
-    contract_api_base: str,
-    ai_platform_api_base: str,
-    *,
-    contract_headers: dict[str, str] | None = None,
-    ai_platform_headers: dict[str, str] | None = None,
-) -> tuple[dict[str, Any], dict[str, Any]]:
-    v1_phase13 = v1_root / "project-docs" / "phase13-legal-knowledge-template-library"
-    v1_templates = v1_phase13 / "contract-templates"
-    v1_clauses = v1_phase13 / "clause-library"
-    v1_rules = v1_phase13 / "review-rules"
-    v1_phase14 = v1_root / "project-docs" / "phase14-legal-fulltext-library"
-    v1_prompt_golden = v1_root / "project-docs" / "phase15-prompt-golden-samples"
-    v1_prompt_assets_file = v1_root / "project-docs" / "phase15-prompt-assets" / "default-prompt-assets.json"
-
-    v2_asset_root = REPO_ROOT / "yuqei-contract-platform" / "services" / "contract-api" / "src" / "yuqei_contract_api" / "template_assets"
-    v2_templates = v2_asset_root / "v1_contract_templates"
-    v2_clauses = v2_asset_root / "v1_clause_library"
-    v2_rules = v2_asset_root / "v1_review_rules"
-    v2_prompt_assets_file = (
-        REPO_ROOT
-        / "yuqei-ai-platform"
-        / "services"
-        / "ai-platform-api"
-        / "src"
-        / "yuqei_ai_platform_api"
-        / "prompt_assets"
-        / "v1_phase15"
-        / "default-prompt-assets.json"
-    )
-    v2_golden_assets = v2_prompt_assets_file.parents[2] / "golden_assets"
-
-    v1_template_ids = asset_ids(v1_templates, "*.md")
-    v2_template_ids = asset_ids(v2_templates, "*.md")
-    v1_clause_ids = asset_ids(v1_clauses, "*.md")
-    v2_clause_ids = asset_ids(v2_clauses, "*.md")
-    v1_rule_ids = asset_ids(v1_rules, "*.md")
-    v2_rule_ids = asset_ids(v2_rules, "*.md")
-    v1_prompt_keys = prompt_keys(read_json_file(v1_prompt_assets_file))
-    v2_prompt_keys = prompt_keys(read_json_file(v2_prompt_assets_file))
-
-    runtime_payloads = {
-        "templates": api_json(contract_api_base, "/api/v1/templates?page_size=100", contract_headers),
-        "clauses": api_json(contract_api_base, "/api/v1/standard-clauses?page_size=100", contract_headers),
-        "rules": api_json(contract_api_base, "/api/v1/review-rules?page_size=100", contract_headers),
-        "legal_articles": api_json(contract_api_base, "/api/v1/legal-articles?page_size=1", contract_headers),
-        "contracts": api_json(contract_api_base, "/api/v1/contracts?page_size=1", contract_headers),
-        "prompts": api_json(ai_platform_api_base, "/api/v1/prompt-templates", ai_platform_headers),
-    }
-    runtime_template_ids = item_ids(runtime_payloads["templates"])
-    runtime_clause_ids = item_ids(runtime_payloads["clauses"])
-    runtime_rule_ids = item_ids(runtime_payloads["rules"])
-    runtime_prompt_items = payload_items(runtime_payloads["prompts"])
-    runtime_prompt_keys = sorted(
-        str(item.get("name"))
-        for item in runtime_prompt_items
-        if item.get("version") == "v1-phase15"
-        and item.get("status") == "published"
-        and item.get("enabled") is True
-        and item.get("name")
-    )
-
-    counts = {
-        "v1_template_variable_schema_files": count_files(v1_templates, "*-variables.json"),
-        "v2_template_variable_schema_files": count_files(v2_templates, "*-variables.json"),
-        "v1_template_markdown_files": len(v1_template_ids),
-        "v2_template_markdown_files": len(v2_template_ids),
-        "v1_clause_library_markdown_files": len(v1_clause_ids),
-        "v2_clause_library_markdown_files": len(v2_clause_ids),
-        "v1_review_rule_markdown_files": len(v1_rule_ids),
-        "v2_review_rule_markdown_files": len(v2_rule_ids),
-        "v1_phase14_legal_files": count_files(v1_phase14, "*"),
-        "v1_prompt_golden_files": count_files(v1_prompt_golden, "*"),
-        "v1_prompt_asset_keys": len(v1_prompt_keys),
-        "v2_packaged_prompt_asset_keys": len(v2_prompt_keys),
-        "v2_ai_golden_asset_files": count_files(v2_golden_assets, "*"),
-        "v2_runtime": {
-            "templates_total": payload_total(runtime_payloads["templates"]),
-            "standard_clauses_total": payload_total(runtime_payloads["clauses"]),
-            "review_rules_total": payload_total(runtime_payloads["rules"]),
-            "legal_articles_total": payload_total(runtime_payloads["legal_articles"]),
-            "contracts_total": payload_total(runtime_payloads["contracts"]),
-            "prompt_templates_total": len(runtime_prompt_items),
-            "v1_phase15_prompt_templates_published": len(runtime_prompt_keys),
-        },
-    }
-    parity = {
-        "templates": compare_ids(v1_template_ids, v2_template_ids, runtime_template_ids),
-        "standard_clauses": compare_ids(v1_clause_ids, v2_clause_ids, runtime_clause_ids),
-        "review_rules": compare_ids(v1_rule_ids, v2_rule_ids, runtime_rule_ids),
-        "prompts": compare_ids(v1_prompt_keys, v2_prompt_keys, runtime_prompt_keys),
-    }
-    return counts, parity
-
-
-def asset_ids(root: Path, pattern: str) -> list[str]:
-    if not root.exists():
-        return []
-    return sorted(
-        path.stem
-        for path in root.rglob(pattern)
-        if path.is_file() and path.name.upper() != "README.MD"
-    )
-
-
-def prompt_keys(payload: Any) -> list[str]:
-    if not isinstance(payload, dict):
-        return []
-    assets = payload.get("assets")
-    if not isinstance(assets, list):
-        return []
-    return sorted(
-        str(item["prompt_key"])
-        for item in assets
-        if isinstance(item, dict) and item.get("prompt_key")
-    )
-
-
-def compare_ids(expected: list[str], packaged: list[str], runtime: list[str]) -> dict[str, Any]:
-    expected_set = set(expected)
-    packaged_set = set(packaged)
-    runtime_set = set(runtime)
-    return {
-        "expected_count": len(expected),
-        "packaged_count": len(packaged),
-        "runtime_count": len(runtime),
-        "missing_in_packaged": sorted(expected_set - packaged_set),
-        "missing_in_runtime": sorted(expected_set - runtime_set),
-        "extra_in_packaged": sorted(packaged_set - expected_set),
-        "extra_in_runtime": sorted(runtime_set - expected_set),
-        "status": "pass" if expected_set <= packaged_set and expected_set <= runtime_set else "fail",
-    }
-
-
-def count_files(root: Path, pattern: str) -> int:
-    if not root.exists():
-        return -1
-    return sum(1 for path in root.rglob(pattern) if path.is_file())
-
-
-def read_json_file(path: Path) -> Any:
-    if not path.exists():
-        return {"__error__": f"Missing file: {path}"}
-    try:
-        return json.loads(path.read_text(encoding="utf-8-sig"))
-    except (OSError, ValueError) as exc:
-        return {"__error__": f"Unable to read {path}: {exc}"}
-
-
-def identity_headers(tenant_id: str, user_id: str) -> dict[str, str]:
-    headers: dict[str, str] = {}
-    if tenant_id.strip():
-        headers["X-Tenant-Id"] = tenant_id.strip()
-    if user_id.strip():
-        headers["X-User-Id"] = user_id.strip()
-    return headers
-
-
-def api_json(base_url: str, path: str, headers: dict[str, str] | None = None) -> Any:
-    try:
-        request = Request(f"{base_url}{path}", headers=headers or {})
-        with urlopen(request, timeout=20) as response:
-            return json.loads(response.read().decode("utf-8"))
-    except (OSError, URLError, TimeoutError, json.JSONDecodeError) as exc:
-        return {"__error__": f"{base_url}{path}: {exc}"}
-
-
-def payload_items(payload: Any) -> list[dict[str, Any]]:
-    if isinstance(payload, list):
-        return [item for item in payload if isinstance(item, dict)]
-    if isinstance(payload, dict) and isinstance(payload.get("items"), list):
-        return [item for item in payload["items"] if isinstance(item, dict)]
-    return []
-
-
-def item_ids(payload: Any) -> list[str]:
-    return sorted(str(item["id"]) for item in payload_items(payload) if item.get("id"))
-
-
-def payload_total(payload: Any) -> int | str:
-    if isinstance(payload, dict) and "__error__" in payload:
-        return str(payload["__error__"])
-    if isinstance(payload, dict) and payload.get("total") is not None:
-        return int(payload["total"])
-    if isinstance(payload, list):
-        return len(payload)
-    return "ERROR: missing total"
-
-
-def derive_status(counts: dict[str, Any], parity: dict[str, Any]) -> str:
-    if any(value == -1 for value in counts.values() if isinstance(value, int)):
-        return "fail"
-    if any(item["status"] == "fail" for item in parity.values()):
-        return "fail"
-    runtime = counts.get("v2_runtime", {})
-    if any(isinstance(value, str) and (value.startswith("ERROR:") or value.startswith("Missing")) for value in runtime.values()):
-        return "warning"
-    if counts["v1_template_variable_schema_files"] != counts["v2_template_variable_schema_files"]:
-        return "warning"
-    return "pass"
-
-
-def render_markdown(result: dict[str, Any]) -> str:
-    counts = result["counts"]
-    runtime = counts["v2_runtime"]
-    rows = [
-        ("V1 template variable schemas", counts["v1_template_variable_schema_files"]),
-        ("V2 template variable schemas", counts["v2_template_variable_schema_files"]),
-        ("V1 template markdown", counts["v1_template_markdown_files"]),
-        ("V2 template markdown", counts["v2_template_markdown_files"]),
-        ("V1 clause markdown", counts["v1_clause_library_markdown_files"]),
-        ("V2 clause markdown", counts["v2_clause_library_markdown_files"]),
-        ("V1 review rule markdown", counts["v1_review_rule_markdown_files"]),
-        ("V2 review rule markdown", counts["v2_review_rule_markdown_files"]),
-        ("V1 Prompt asset keys", counts["v1_prompt_asset_keys"]),
-        ("V2 packaged Prompt asset keys", counts["v2_packaged_prompt_asset_keys"]),
-        ("V1 Prompt golden files", counts["v1_prompt_golden_files"]),
-        ("V2 AI golden asset files", counts["v2_ai_golden_asset_files"]),
-        ("V2 runtime templates", runtime["templates_total"]),
-        ("V2 runtime standard clauses", runtime["standard_clauses_total"]),
-        ("V2 runtime review rules", runtime["review_rules_total"]),
-        ("V2 runtime legal articles", runtime["legal_articles_total"]),
-        ("V2 runtime Prompt templates", runtime["prompt_templates_total"]),
-        ("V2 published V1 Phase15 Prompts", runtime["v1_phase15_prompt_templates_published"]),
-        ("V2 runtime contracts", runtime["contracts_total"]),
-    ]
-    lines = [
-        "# V2 Asset Count",
-        "",
-        f"Status: `{result['status']}`",
-        "",
-        f"Checked at: `{result['checked_at']}`",
-        "",
-        "| Asset | Count |",
-        "| --- | ---: |",
-    ]
-    lines.extend(f"| {name} | {value} |" for name, value in rows)
-    lines.extend(["", "## ID Parity", "", "| Asset | Expected | Packaged | Runtime | Status |", "| --- | ---: | ---: | ---: | --- |"])
-    for name, item in result["parity"].items():
-        lines.append(
-            f"| {name} | {item['expected_count']} | {item['packaged_count']} | {item['runtime_count']} | {item['status']} |"
-        )
-        if item["missing_in_packaged"]:
-            lines.append(f"| {name} missing packaged | - | `{', '.join(item['missing_in_packaged'])}` | - | fail |")
-        if item["missing_in_runtime"]:
-            lines.append(f"| {name} missing runtime | - | - | `{', '.join(item['missing_in_runtime'])}` | fail |")
-    lines.extend(["", "## Notes", ""])
-    lines.extend(f"- {note}" for note in result["notes"])
-    lines.append("")
-    return "\n".join(lines)
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

yuqei-infra/scripts/deploy-v2-smoke-docker.ps1

@@ -449,7 +449,7 @@ docker compose -p $(ConvertTo-ShellSingleQuote $ComposeProjectName) --env-file $
     Add-Check "ai_golden_sample_gate" "AI golden sample asset gate" "fail" "Golden sample asset gate failed." $true @{ log = $goldenGate.log }
     throw "AI golden sample asset gate failed. See $($goldenGate.log)"
   }
-  Add-Check "ai_golden_sample_gate" "AI golden sample asset gate" "pass" "V1 golden assets passed: 90 samples, 6 suites, 16 checks." $true @{ log = $goldenGate.log }
+  Add-Check "ai_golden_sample_gate" "AI golden sample asset gate" "pass" "Packaged golden assets passed: 90 samples, 6 suites, 16 checks." $true @{ log = $goldenGate.log }
 
   $dbCheck = Invoke-Remote "remote-db-check" @"
 set -euo pipefail

yuqei-infra/scripts/inspect-v2-database-state.ps1

@@ -1,47 +0,0 @@
-param(
-    [string]$HostName = "192.168.3.90",
-    [string]$User = "root",
-    [int]$SshPort = 22
-)
-
-Set-StrictMode -Version Latest
-$ErrorActionPreference = "Stop"
-
-$remoteScript = @'
-set -eu
-
-echo "## host-postgres-databases"
-sudo -u postgres psql -Atc "SELECT datname FROM pg_database WHERE datistemplate = false ORDER BY datname;"
-
-echo
-echo "## host-postgres-summaries"
-for db in ai_contract_platform yuqei_contract yuqei_ai yuqei_ops yuqei_v2 postgres; do
-  exists=$(sudo -u postgres psql -Atc "SELECT 1 FROM pg_database WHERE datname = '$db';")
-  if [ "$exists" = "1" ]; then
-    tables=$(sudo -u postgres psql -d "$db" -Atc "SELECT count(*) FROM information_schema.tables WHERE table_schema='public' AND table_type='BASE TABLE';")
-    versions=$(sudo -u postgres psql -d "$db" -Atc "SELECT table_name FROM information_schema.tables WHERE table_schema='public' AND table_name LIKE '%alembic_version%' ORDER BY table_name;" | paste -sd ',')
-    heads=""
-    for vt in alembic_version contract_api_alembic_version ai_platform_alembic_version; do
-      has=$(sudo -u postgres psql -d "$db" -Atc "SELECT 1 FROM information_schema.tables WHERE table_schema='public' AND table_name='$vt';")
-      if [ "$has" = "1" ]; then
-        val=$(sudo -u postgres psql -d "$db" -Atc "SELECT version_num FROM $vt LIMIT 5;" | paste -sd ',')
-        heads="$heads $vt=$val"
-      fi
-    done
-    markers=$(sudo -u postgres psql -d "$db" -Atc "SELECT table_name FROM information_schema.tables WHERE table_schema='public' AND table_name IN ('contracts','generated_contracts','knowledge_documents','ai_run_audits','legal_change_impact_scans') ORDER BY table_name;" | paste -sd ',')
-    echo "$db|tables=$tables|versions=$versions|heads=$heads|markers=$markers"
-  else
-    echo "$db|missing"
-  fi
-done
-
-echo
-echo "## docker-v2-smoke-containers"
-docker ps -a --format '{{.Names}}|{{.Status}}|{{.Ports}}' | grep -E 'yuqei-contract-v2-smoke|ai-contract-platform|postgres' || true
-
-echo
-echo "## docker-v2-smoke-volumes"
-docker volume ls --format '{{.Name}}' | grep -E 'yuqei-contract-v2-smoke|ai-contract-platform|postgres' || true
-'@
-
-$remoteScript | ssh -p $SshPort "$User@$HostName" "bash -s"

yuqei-infra/scripts/run-ai-provider-acceptance.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-"""Run real-provider legal QA and contract-review acceptance against V1 golden samples."""
+"""Run real-provider legal QA and contract-review acceptance against packaged golden samples."""
 
 from __future__ import annotations
 

yuqei-infra/tests/test_v2_only_repository_layout.py

@@ -0,0 +1,58 @@
+from pathlib import Path
+
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+
+RETIRED_PRODUCT_ROOTS = (
+    "backend",
+    "backups",
+    "database",
+    "deploy",
+    "frontend",
+    "infra",
+    "mobile",
+    "scripts",
+    "temp-manual-laws",
+)
+
+RETIRED_V1_ARTIFACTS = (
+    "project-docs/V2-vs-V1-feature-gap-audit-2026-07-03.md",
+    "yuqei-ai-platform/services/ai-platform-api/src/yuqei_ai_platform_api/v1_prompt_assets_import.py",
+    "yuqei-ai-platform/services/ai-platform-api/tests/test_v1_prompt_assets_import.py",
+    "yuqei-contract-platform/services/contract-api/src/yuqei_contract_api/v1_legal_fulltext_import.py",
+    "yuqei-contract-platform/services/contract-api/tests/test_v1_legal_fulltext_import.py",
+    "yuqei-infra/scripts/compare-v1-v2-assets.py",
+    "yuqei-infra/scripts/inspect-v2-database-state.ps1",
+)
+
+
+def test_retired_product_roots_are_not_present() -> None:
+    present = [name for name in RETIRED_PRODUCT_ROOTS if (REPO_ROOT / name).exists()]
+
+    assert present == []
+
+
+def test_retired_v1_only_artifacts_are_not_present() -> None:
+    present = [path for path in RETIRED_V1_ARTIFACTS if (REPO_ROOT / path).exists()]
+
+    assert present == []
+
+
+def test_repository_entrypoints_describe_a_single_v2_product_line() -> None:
+    entrypoint_text = "\n".join(
+        (REPO_ROOT / path).read_text(encoding="utf-8")
+        for path in ("README.md", "V2-README.md")
+    )
+
+    assert "previous complete V1 implementation" not in entrypoint_text
+    assert "previous V1 implementation" not in entrypoint_text
+    assert "`V1` branch" not in entrypoint_text
+
+
+def test_active_pages_do_not_present_retired_product_assets() -> None:
+    app_root = REPO_ROOT / "yuqei-contract-platform/apps/contract-web/src/app"
+    page_text = "\n".join(path.read_text(encoding="utf-8") for path in app_root.rglob("*.tsx"))
+
+    assert "V1 Phase" not in page_text
+    assert "V1 模板资产" not in page_text
+    assert "V1 全量" not in page_text