admin/danyang-match
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
106 Commits 2 Branches 8 Tags
c77e6fcd6383d81db793d09565c54aade7896baa
Commit Graph
33 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
WorkBuddy c77e6fcd63 chore(docs): 删除已被复验推翻或取代的失效历史文档（65 份） 
- V1.0 自报式 Phase-XX-Completion-Report 26 份（README 已声明其不可信）
- V1.1 首轮 Phase-XX-V11-Completion-Report 21 份（已被对应 Recovery 版复验推翻/覆盖）
- V1.0 时代审查/发布报告 13 份（CODE_REVIEW/FIX_PLAN/ISSUE_TRACKER/发布与 Go-NoGo/UAT/渗透测试/P0 指令/REAUDIT/两次评估）
- 自declared被取代的旧计划 4 份（RECOVERY_PLAN_2026-09-09、当前进度校正、09-14 两份 superpowers 计划）
- IDEA.md（一行备忘）

引用核查：代码/脚本/CI/配置零引用；STATUS.md 与 REMAINING_WORK.md 中两处现行引用随后续文档同步修正；
Phase-10 Recovery 报告与 2026-09-13 UI spec 中的历史叙述引用按原文保留。
治理详情见 docs/文档治理报告-2026-09-18.md。
 2026-09-18 20:23:43 +08:00
WorkBuddy 3ff0c0641a chore(admin): 生产管理端端口 3000 → 13520；守卫扩至三处并容错 CI 
生产 admin 容器发布端口原为 127.0.0.1:3000 —— 3000 是本机最易被其他项目
（videoos / ai-callcenter）抢用的端口，与新定的管理端端口 13520 不一致。
已改为 127.0.0.1:13520:80（仍绑定 loopback，对外访问走 nginx 80/443 + /admin/）。

⚠️ 该实体改动无法进入本次提交：infrastructure/docker-compose.prod.yml 位于
.gitignore 的「Secrets and certificates」段，刻意不进版本控制（含按主机定制的
部署参数）。因此本次提交承载的是「决策记录 + 守卫」，端口改动只在部署机生效。

守卫 ui-tests/ui-document-baseline-contracts.mjs：
- 由「两处一致」扩为「三处一致」：admin-web/vite.config.ts（权威声明）、
  docs/STATUS.md（给人照做的）、docker-compose.prod.yml（生产发布端口）。
- 生产 compose 采取「存在才断言」：CI 全新克隆上该文件不存在，硬 read() 会抛
  ENOENT 让门控假红。已实测——移走文件 → 打印 skip 且 EXIT=0；放回 → 断言恢复。
- 明确排除另外两处管理端入口并写清理由：本地整套栈的管理端经 nginx 6688 暴露
  （6688 是 nginx 入口，不是纯管理端端口）；staging 的 3001 是为与生产同机共存
  而按「生产 + 1」错开（5433/6380/8081/3001），且被 tests/uat/run-uat.sh 的
  UAT_ADMIN_URL 默认值依赖 —— 改动它们要同步消费方，不该由本守卫断言等于 vite 端口。

docs/STATUS.md：新增「管理端端口口径」表，记录四种管理端入口的端口与取舍理由，
并显式标注生产 compose 不在版本控制内（否则后人会以为改了没生效）。

验证：ui-tests 门控 14/14；负向验证 prod compose 改回 3000 → 精准报
「docker-compose.prod.yml=3000，vite.config.ts=13520」；文件缺失 → skip 且不假红；
还原后 md5 逐字节一致（71a30653…）；docker compose config 解析出
published=13520 且 host_ip=127.0.0.1（loopback 绑定未丢失）。
 2026-09-18 19:49:50 +08:00
WorkBuddy 9c26d4b7b2 chore(admin-web): 开发端口 3000 → 13520，并禁止端口静默漂移 
本机长期并行运行三个项目（videoos / ai-callcenter / 本项目），3000 属最易被
抢用的端口之一。按用户决定改为 13520。

- admin-web/vite.config.ts：server.port 3000 → 13520
- 同处新增 strictPort: true。vite 默认在端口被占用时静默 +1 退到下一个可用端口，
  结果是「文档写 13520、实际跑在 13521」且全程不报错 —— 与本项目反复处理的
  「声明与兑现失联」同族。加上后端口被占用会直接失败。
- docs/STATUS.md「当前运行状态」同步为 13520。
- ui-tests/ui-document-baseline-contracts.mjs：新增端口一致性守卫。

  端口写在两处（vite.config.ts 是权威，STATUS.md 给人照做），只改一处不会让任何
  测试变红，却会让照文档操作的人打不开页面。守卫断言的是「两处一致」而非写死
  13520 —— 允许改端口，允许的是「一起改」，不是「只改一处」。
  两处解析均用「解析失败即失败」写法，避免正则匹配不到时断言退化成与 undefined
  比较而失去意义（本项目踩过的守卫静默失效坑）。
  用 assert.equal 而非 assert.match：match 失败会把整份 STATUS.md（3000+ 行）
  打进报错，噪声会盖掉真正的不一致信息 —— 负向验证时实测确认。

- .gitignore：忽略 admin-web/vite.config.ts.timestamp-*.mjs。
  vite 在 dev server 被强杀（未优雅退出）时遗留该配置临时文件，本机已多次出现。

未改动（有意）：
- CORS 不受影响。CorsConfig 默认拒绝所有跨域是刻意设计，且注释已写明「管理端走
  Vite 代理（同源）」—— dev 端口变化不产生跨域。
- infrastructure/docker-compose.prod.yml 的 127.0.0.1:3000:80 是生产 admin 容器
  的发布端口，与开发服务器无关，未动。
- releases/v1.0.0-RC.* 内的 3000 为历史发布快照，属冻结记录，不改写。

验证
- UI 契约门控 14/14 全绿（含新增守卫）；vue-tsc --noEmit EXIT=0。
- 负向验证：把 STATUS.md 改回 3000 → 守卫变红并精确报出
  「管理后台端口不一致：STATUS.md=3000，vite.config.ts=13520」及行号；还原后复绿。
- 真实启停：dev server 实跑 http://localhost:13520/（HTTP 200），旧端口 3000 无监听。
- strictPort 负向验证：端口被占用时第二实例立即以
  「Error: Port 13520 is already in use」退出，未漂移到 13521。
- vite build EXIT=0（19.35s）。

注：vite build 直接执行会失败，栈停在 emptyDir(dist) 且全部位于本机
node-safe-delete-shim（批量删除守卫）内，与本次改动无关；改用全新 outDir 复跑即
成功，据此排除「改动导致构建失败」。该现象此前无记录，已写入技能备查。
 2026-09-18 19:36:07 +08:00
WorkBuddy c561bf46af docs(idempotency): 统一报告中的计数口径（25 处调用点 / 30 处声明）并补 CI E2E 缺陷到结论行 2026-09-18 16:40:09 +08:00
WorkBuddy 2766e53fe2 docs: 幂等键检修报告；勘误 API_INVENTORY 与 HTTP 报告；同步 STATUS/JIRA 
新增 docs/幂等键检修报告-2026-09-18.md：记录双向断裂（后端零处读取该头、
前端 Date.now 每次生成新键）、五处审计发现、机制设计与五个取舍
（先占位后执行 / 重放 / 失败释放占位 / 孤儿占位接管 / 作用域含 actor_id）、
列类型不变量扫描、E2E 与守门证据、三处同形状问题的连带修复。

同轮勘误：

- docs/HTTP语义检修报告-2026-09-18.md：其中「Idempotency-Key 已支持」一句
  在写下时不成立（后端当时零处读取），已显式标注勘误并保留原文，
  以免「结论先于核实」被当作既成事实继续引用。
- docs/api/API_INVENTORY.md：原记载 POST /activities/{id}/interests
  「提交活动互选」，后端从未实现（Phase 16 已并入 /feedback 的
  interestTargetUserIds），按真实端点重列并附勘误说明。

docs/STATUS.md：迁移范围更新为 V001~V046；回归行更新为 669 / 0 / 0 / 3；
新增「幂等键检修」章节；补登 HTTP 语义检修章节在两份文档中的缺失条目。
docs/JIRA_BACKLOG.md：补登 HTTP 语义检修与幂等键检修两行
（此前均未登记）。
 2026-09-18 16:37:47 +08:00
WorkBuddy af5412646b test(e2e): 运营登录只用 seed 账号；补幂等 E2E 与两处机械守卫 
1) 修复一处从未暴露的 CI 缺陷

库中 operator 账号没有任何迁移种子，DemoDataSeeder 也不创建运营账号
（只统计数量），全新库唯一的账号来源是 seed_e2e_operators.sql。
而 conftest 的 admin_token 写死 username="admin" —— 那是某台开发机上
手工创建的遗留账号：本机有它所以本地一直跑绿，CI 全新库没有它 →
该 fixture 必然 401 → e2e job 从未真正通过过。
同一文件里「外加迁移 seed 的 admin」这句注释也是纯粹的虚构。

改用 seed 创建的 e2e_admin（已绑定 super_admin，语义等价）。

2) fixture_account_guard_test.py（2 用例）

机械拦住「fixture 依赖本机手工数据」：从 seed SQL 解析账号清单，
扫描 conftest 的所有登录用户名，任何一个不在 seed 内即变红；
另一条锁住推理前提（迁移里确实没有 operator 种子）。
两条都带「解析失败即失败」断言 —— 守卫自身静默失效是本项目反复踩过的坑。

3) idempotency_test.py（7 用例）

同键重放响应逐字节相同、同键重复提交只产生一条记录、body 字段幂等未回归、
不同键相互独立、同键作用域按 actor 隔离、5 线程并发同键业务恰好落库一次、
无键重复 Like 不得返回 500。

4) conftest.ApiClient.call 支持 headers 覆盖，供幂等用例携带自定义键。
 2026-09-18 16:37:33 +08:00
WorkBuddy 8dee67040c fix(miniapp): 幂等键改为按操作稳定；删除指向 404 的互选死代码 
1) 幂等键（配合后端机制）

原写法把 Date.now() 拼进幂等键，每次调用都生成新键，即使后端实现了幂等
也永远不会命中 —— 声明端与实现端各自都坏。

api.idempotencyKey(scope, subject)：键缓存在 _pendingIdempotency（TTL 10min），
同一逻辑操作在成功之前复用同一把键，收到 200+success 才释放；失败（含网络错误）
不释放，下次重试沿用同一把键。25 处调用点（8 个 service + 2 个 TS 页面）全部替换，
Date.now 幂等键写法残留经 grep 为 0。

2) 删除死代码 submitInterests

它请求 POST /activities/{id}/interests —— 后端从未实现该端点
（Phase 16 已把表达意向合并进 /activities/{id}/feedback 的
interestTargetUserIds）。全仓确认无任何页面/组件调用方。
两个文件各留注释说明删除原因与恢复条件：若产品确认需要独立的
「提交互选」端点，应先定义语义并补后端实现，再恢复前端方法。

store 头部关于 AGENTS 第 6 条（单向意向保密）的注释同步改为引用
submitFeedback 返回的 newMutuals。
 2026-09-18 16:37:33 +08:00
WorkBuddy 69cf6a2cc7 feat(idempotency): 为 30 个写端点实现 Idempotency-Key 
规范第八章要求 7 类写端点支持 Idempotency-Key，第 146 章验收口径为
「最终不能生成重复核心数据」。检修前的实际状态：后端零处读取该头，
小程序 25 处调用点发送全部石沉大海。

一处实现、零业务代码改动、逐端点声明：
- V045 新建 idempotency_records + 唯一索引 uk_idempotency_scope
- V046 修正 actor_id 列类型 VARCHAR -> uuid（见下）
- IdempotencyInterceptor：先占位后执行 / 重放首次响应 / 处理中回 409 /
  孤儿占位 300s 可接管 / 失败释放占位
- IdempotencyResponseCacheFilter：缓存响应体供逐字节重放
- WebMvcConfig：本项目第一个通用 WebMvcConfigurer
- @Idempotent 声明 30 处（含补 CheckinController 手工签到的漏标）
- GlobalExceptionHandler：IllegalStateException（30 处业务状态冲突）与
  DuplicateKeyException（并发重复提交的最后防线）此前均无映射 → 落 500，
  而 500 的语义是「可重试」，调用方于是重试并制造更多重复。现映射为 409；
  刻意不映射父类 DataIntegrityViolationException（CHECK 违反应由 500 暴露）。
- schema-h2.sql 补 idempotency_records

V046 的必要性：项目全局注册的 UUIDStringTypeHandler 对任何合法 UUID 形式的
String 参数以显式 uuid 类型发送，而 SQL 侧 stringtype=unspecified 只影响以
unknown 类型发送的参数，覆盖不到它。因此真正的不变量不是「JDBC 参数配好了」，
而是「列类型必须是 uuid」。V045 初版把 actor_id 建成 VARCHAR(64) 时，
带幂等键的请求在真实 PG 上全部 500。

全库扫描确认该不变量已收敛：所有 *_id VARCHAR 列中 actor_id 是唯一违规，
client_message_id / provider_request_id 为外部 id，
resource_id / source_id 在 Java 侧零处作查询条件。
 2026-09-18 16:37:11 +08:00
WorkBuddy ed725daccd docs(status): record identity scope convergence and the PathPattern ** trap 2026-09-18 15:06:06 +08:00
WorkBuddy cbb4f64fdc fix(security): stop operator identity from reaching user endpoints; guard path rules 
Continues the HTTP-semantics work. Fixing 401/403 exposed a worse defect on the
same axis: who is allowed to reach which endpoint.

Root cause chain:
- PlatformPrincipal has two implementations (OperatorPrincipal / UserPrincipal),
  chosen by JwtAuthFilter from the token type.
- `@AuthenticationPrincipal UserPrincipal` against an OperatorPrincipal is a type
  mismatch, and Spring Security returns null (errorOnInvalidType defaults to
  false) instead of failing.
- 244 uses of that annotation repo-wide; only 48 null-check. The rest NPE -> 500
  when reached with an operator identity.
- The fallback rule was `.authenticated()`, which an operator token satisfies, so
  such requests did reach user-facing controllers.

Adding ~196 null checks would be mechanical, easy to miss, and does nothing to
constrain future code. The mismatch is between identity type and endpoint, so it
is fixed once, at the authentication layer: the fallback becomes
`.hasRole("USER")` (operator -> ROLE_OPERATOR, user -> ROLE_USER, per JwtAuthFilter).
Operator requests to user endpoints now get 403 before any controller runs.

Trap worth recording: operator endpoints are not all under /operator/ —
/counseling/services/operator/queue, /content/operator/articles,
/activities/operator/checkin/scan, /service-appointments/operator/list and others
sit deeper inside business paths. The first attempt covered them with
`requestMatchers("/**/operator/**")`, which fails at runtime:

  PatternParseException: No more pattern data allowed after {*...} or ** pattern element

Spring's PathPattern permits `**` only as the final element. Worse, the symptom is
a 500 from the whole filter chain (Tomcat HTML error page, not our JSON), and E2E
went from 37 passed to 33 failed — an error far removed from its cause. Now the
nine prefixes are listed explicitly, each ending in /**.

Because explicit enumeration can be forgotten as endpoints are added, and because
the failure mode is "operators locked out of their own admin module" (invisible to
user-side tests), SecurityPathCoverageTest parses the OPERATOR rule prefixes out of
SecurityConfig source, walks every *Controller.java, joins class-level and
method-level mapping paths, and asserts every path containing /operator/ is
covered (/auth/operator/** excluded by design). Two things that would silently
neuter it, both handled: it must parse only the .hasRole("OPERATOR") chain — pulling
in the permitAll or fallback rules would mark everything covered — and the test
working directory is the repo root, not the Maven module root.

Verification: negatively validated by removing /content/operator/** -> exactly one
test turns red listing the six affected ContentController endpoints; restored.
E2E 40 passed (37 + 3 identity-isolation cases). Backend regression 664 / 0
failures / 0 errors / 3 skipped.
 2026-09-18 15:05:41 +08:00
WorkBuddy 3a4e03a9f9 docs(status): record HTTP semantics remediation; note 3 test assertions that encoded old wrong behavior 2026-09-18 14:38:59 +08:00
WorkBuddy 7682abd85c fix(api): give failures real HTTP semantics; fix 401/403 confusion and CORS default 
Phase 23B ended with an E2E assertion changed to expect "HTTP 200 + body.code
== BAD_REQUEST", documented at the time as "this project's convention". That
description was accurate about the code and wrong about the intent: it described
a defect, and I made the test accommodate it instead of fixing it.

Auditing it found the convention was already broken, not merely risky:

- ~100 sites (105, across 30 controllers) do `return ApiResponse.error(...)` /
  `forbidden()` / `notFound()` with no exception and no `@ResponseStatus`, so
  the HTTP status is always 200. Gateways and APM cannot separate 4xx from 5xx;
  callers cannot tell "fix your request" (400) from "don't retry" (403) from
  "check state first" (409) from "retry later" (429).
- miniapp/utils/api.js already had 401/403/429 branches that could never be
  reached, since the backend never returned those codes. The concrete damage:
  an expired token returned 403, so `onSessionExpired` never fired and users
  were stuck with no way back to the login flow.
- SecurityConfig had no `authenticationEntryPoint`, so Spring's default
  `Http403ForbiddenEntryPoint` returned 403 for unauthenticated calls, with a
  framework body that has no `code` field at all.
- CorsConfig defaulted `allowed-origins` to `*` together with
  `allowCredentials(true)`, while its comment claimed "localhost only by
  default". Comment contradicting code is worse than no comment: it reads as
  safe.
- Only 4 files / 15 assertions in the whole backend suite touched HTTP status,
  which is how this survived.

Approach: one definition, two applications, zero controller changes.
- ApiResponseCode gains an `httpStatus` field — the single source of truth. The
  constructor requires it, so a new code cannot be added without one.
- GlobalExceptionHandler reads status from the enum instead of hardcoding
  `@ResponseStatus` per handler.
- New ApiResponseStatusAdvice (ResponseBodyAdvice) fills in the status just
  before the body is written, covering all ~100 direct-return sites without
  touching them. Two constraints, both noted in code: only downgrade when the
  current status is still 2xx (the advice also sees exception-handler results,
  and must not override an already-correct 401/403/409), and never alter the
  body.
- SecurityConfig: entry point returns 401 with a uniform ApiResponse body;
  access denied handler returns 403 likewise.
- CorsConfig: default changed to empty (deny). Both clients are same-origin
  anyway — nginx proxies /api/ and Vite proxies /api — so nothing depends on it.
- miniapp api.js: 401/403/429/409 branches now prefer the server's
  `body.message`, otherwise "账号已被封禁" would be flattened to "无权限访问".

Implementation trap worth remembering: `ServerHttpResponse` has `setStatusCode`
but no getter, so reading the current status requires casting to
`ServletServerHttpResponse`.

Verification: 37 E2E passed (33 + 4 new contract cases), including a live-stack
check that a controller's direct `notFound()` now yields 404 — that path throws
no exception and never reaches the exception handler, so if the advice were
inactive it would silently return 200 while the enum unit tests stayed green.
Backend regression 662 / 0 failures / 0 errors / 3 skipped (+6 new guards).

The first regression run had 3 failures, all tests asserting the old wrong
behavior (anonymous access expected 403; outsiders reading a conversation
expected 200). Fixed the assertions, not the implementation. Where 404 is used
instead of 403 for "invisible", that is deliberate — it prevents probing whether
another user's conversation exists — and the substantive assertions (e.g. that an
unauthorized response must not leak message bodies) were kept.

Not done, needs a product decision: third-party credential issuance, whether
rate limiting is actually active, publishing OpenAPI externally, per-caller
audit and quotas, pagination/idempotency coverage, desensitization boundaries.
 2026-09-18 14:34:58 +08:00
WorkBuddy 730dddde30 test(e2e): Phase 23B — rebuild E2E infrastructure; fix 6 real prod defects it exposed 
Phase 23 closed as PARTIAL: fake coverage was gone, but the E2E layer itself
was still empty — no conftest.py anywhere, ~15 fixtures that never existed,
6 files crashing at import on os.environ["E2E_PASSWORD"], and CI never ran E2E.
STATUS's "10 E2E files / ~80+ cases ✅" described files that had never executed.

Infrastructure delivered:
- tests/e2e/conftest.py: ApiClient (trust_env=False — a host http_proxy turns
  localhost calls into 502), operator_token(username) / new_user / admin_token
  fixtures, and a session probe that calls pytest.exit(4) when the stack is
  unreachable. NOT skip: skip would launder "did not run" into "passed".
- pytest.ini, tests/e2e/requirements.txt, and tests/e2e/sql/
  seed_e2e_operators.sql — migrations contain NO operator seed at all, so a
  fresh database has zero logins; the seed creates 6 e2e accounts with role
  bindings, idempotent, password injected via `psql -v pwd`.
- CI job `e2e`: compose up -> poll /health/live (timeout = fail) -> seed ->
  pytest -> always teardown. Gate D-23-D is now closed.
- 8 E2E files migrated off import-time env reads; tautological assertions
  (`x == t or not x`), a counselor token standing in for content_editor, and a
  bare `status_code == 200` claiming to prove INV-020 were replaced with real
  permission/behavior assertions.

Six real production defects found only by running against real PostgreSQL
(H2 has zero coverage for all of them):
- D-23B-A (Critical): stale V042__add_appointment_operator_register_permission.sql
  left in target/classes; src had replaced it with V042__add_service_location_
  announcements.sql. Packing without clean put two V042 in the jar -> Flyway
  "Found more than one migration with version 042" -> app would not start.
  Guard added: testBuildOutputHasNoDuplicateMigrationVersions scans build output
  (src is clean, so only this catches it).
- D-23B-B (High): non-UUID path variable reached the DB -> 500 + leaked
  `invalid input syntax for type uuid`. Now rejected at the controller.
- D-23B-C (High): 6 E2E files raised KeyError at import -> never collected.
- D-23B-E (Critical): permissions = 48, role_permissions for super_admin = 18,
  matchmaker = 0. V015's "grant all to super_admin" was a point-in-time snapshot;
  the 30 permissions added later were never granted. On a fresh stack an admin
  cannot create a service station and matchmakers can do nothing. V044 restores
  super_admin to its documented definition (not an expansion); the matchmaker
  permission set needs a product decision and is left as a residual.
- D-23B-I (Medium): IllegalArgumentException (e.g. invalid activity type) was
  unmapped -> 500. Now BAD_REQUEST.
- D-23B-J (High): `operator does not exist: uuid = character varying` — Java
  ids are String, columns are uuid, PostgreSQL does not implicitly cast in
  comparisons. Fixed at the source with JDBC `stringtype=unspecified` (docker
  profile); production datasource comes from deploy env and needs the same.

One judgement retracted: an initial read suggested service_appointment.register
was missing. It is not — the endpoint uses service.assisted_registration, seeded
in V029, and the code deliberately avoids a duplicate permission. Nothing was
changed on that basis.

Verification: 33 E2E passed against a live stack; unreachable stack and a
missing operator password both exit 4 (never skip/pass); backend regression
656 / 0 failures / 0 errors / 3 skipped (baseline 655, +1 guard, negatively
validated by injecting a duplicate V042 -> exactly one test turned red).
 2026-09-18 13:50:19 +08:00
WorkBuddy ae1bd2b734 test(phase23): KEEP-VERIFY — remove 44 pass shells + fake INV tests, add shell guard 
Phase 23 "可信系统测试重建 (REFACTOR)" re-audit. STATUS claimed "10 E2E
files / ~80+ cases" and "INV-011~020 ✅"; measurement showed the opposite.

D-23-A (Critical): V11InvariantsTest's 14 cases asserted only Constants
strings. INV-016 "Banner target types" actually asserted ARTICLE_STATUS_*;
INV-017 "home 8 entries" declared local arrays then asserted
firstRow.length == 4 (tautology, never called the home service);
INV-018/019/020 used appointment/article/account status constants as
stand-ins for permission and data-scope checks. Rewritten to 12 cases:
4 real service-behavior guards (INV-013 via real CounselingService.getDetail
— Matchmaker sees null userRequestSummary, with an assigned-Counselor
counter-proof so global sanitization cannot fake a pass; INV-016 reads the
production ALLOWED_TARGET_TYPES whitelist and asserts no external/http/url/
web entry; INV-017 calls real getHome and pins all 8 codes + order;
INV-019 asserts content_editor reads null), plus assertBehaviorTestExists
turning "行为验证见 XxxTest" pointers into executable Class.forName contracts.

D-23-B (High): 44 `pass` shells across 6 files in tests/e2e (4 files were
100% shells). Gate forbids skip/pass shells. Rewritten in place as honest
placeholders with the full scenario lists preserved.

D-23-C (High, residual): the pytest E2E suite is not executable at all —
no conftest.py anywhere, ~15 fixtures (api_client/verified_user/
banner_factory/...) do not exist. Only run_v11_tests.py really runs
(16 cases, needs a live stack + E2E_PASSWORD).

D-23-D (High, residual, NOT fixed): CI does not chain E2E as the plan
requires. Needs an infrastructure decision on how CI provisions a live
stack; not implemented rather than written unverified.

D-23-E (Medium): the CI migration job claimed to "验证 migration SQL 语法"
but only checked file non-emptiness, while the real syntax/baseline checks
live in MigrationValidationTest (already run by the backend job). Replaced
with a real sequence check (naming / non-empty / unique versions / no
gaps); verified by extracting the exact inline script and running it.

New gate: scripts/check-no-empty-tests.py (AST scan, blocks pass-only and
unconditionally skipped tests) + a CI test-shell-guard job. Verified both
ways: 0 findings on the real tree, 2 precise findings on injected shells.

Negative validation: 3 injected production defects (external_url in the
banner whitelist, swapped home entry order, isMatchmaker returning the full
record) turned exactly inv016/inv017/inv013 red, 9 others green, zero
unrelated failures; md5sum -c confirmed byte-identical restore.

Full regression: 655 tests / 652 passed / 0 failures / 0 errors / 3 skipped
(baseline 657; -2 because 14 vacuous cases were merged into 12). The 3
skips carry explicit reasons (2x @EnabledIfEnvironmentVariable(CI=true),
1x @Disabled "需要完整数据库连接") — none is a shell.

Phase 23 is PARTIALLY closed: fake coverage is gone, but the UAT gate
requires the end-to-end chain to actually execute, so "失败为零可进入 UAT"
is NOT claimed.
 2026-09-18 11:49:18 +08:00
WorkBuddy 050d005c0a docs(status): correct two stale duplicate Phase 22 ✅ tables disproved by KEEP-VERIFY 
STATUS.md contained two duplicate legacy Phase 22 blocks still claiming:
- "application.yml 默认配置 ✅ (15个V1.1 flag)" / "application-dev.yml +
  application-docker.yml 配置 ✅" — verified false: all four yml files
  (application.yml / -dev / -docker / -production) contain neither a `features:`
  block nor any flag key; defaults are Java field defaults only.
- "P22-T01..T05 ✅" and "P22-T07 ✅" — these are dead config: zero business
  call sites for the five module toggles, zero references for the private
  storage bucket flags.

Both blocks now carry a "superseded by the 2026-09-18 authoritative section"
notice with per-row measured verdicts (T06 remains ✅ as genuinely enforced).
No code change; doc-only.
 2026-09-18 11:08:12 +08:00
WorkBuddy 2fbdbc32d0 fix(feature-flags): close Phase 22 KEEP-VERIFY — 2 real defects + doc overclaims 
KEEP-VERIFY re-audit of the Feature Flag configuration extension (STATUS's
"Phase 22 部署运维配置扩展"): marked ✅ but contained real code defects and
overstated completion.

Real defects fixed:
- D-22-A (High): FeatureFlagController.updateFlag used reflection
  `getMethod("set"+cap, boolean.class)` and only supported boolean flags.
  7 non-boolean flags (livenessProvider, localEvidenceAllowedTypes,
  counselingAttachmentsBucket, localEvidenceMaxFiles, localEvidenceMaxFileSizeMB,
  localEvidenceRetentionDays) failed to update via the API. Now dispatches by
  field type (boolean/int/String).
- D-22-B (Medium): `body.getOrDefault("value", false)` silently set false when
  value was missing and crashed on non-boolean casts. Now returns 400 on missing
  value / type mismatch and leaves the original value unchanged.

Doc overclaims corrected (no silent scope creep):
- T01-T05 module-toggle flags (isCounselingEnabled etc.) have zero business
  call sites -> dead config (exposed but not gating). Left for each module's
  own phase to wire; documented as residual.
- T07 private-storage bucket flags have zero references; storage already uses
  object-storage.bucket-private. Documented as residual.
- STATUS claimed "application.yml defaults (15 flags)"; yml has no `features:`
  block, defaults are Java field defaults. Functionally fine, docs corrected.

T06 upload policy verified actually enforced in LocalEligibilityService +
PhotoService (kept ✅).

Guard tests: new FeatureFlagControllerTest (7 cases: 21 keys exposed, boolean/
string/int update reflected, unknown flag / missing value / wrong type -> 400).
Negative validation: reverting to boolean-only dispatch made exactly 2 update
cases fail red, then green after restore — proves guards are real.

Full regression: 657 tests / 0 failures / 0 errors / 3 skipped (baseline 650, +7).
 2026-09-18 11:00:08 +08:00
WorkBuddy 0580da0c62 fix(account-lifecycle): close Phase 21 KEEP-VERIFY — 6 real defects + fake-green tests 
KEEP-VERIFY re-audit of account lifecycle (EXTEND): STATUS marked ✅ but
contained hard Gate violations and dead code.

Real defects fixed:
- D-21-A: AuthService.checkAccountStatus did not reject CLOSED → closed
  account could still log in (Gate ❌→✅); added CLOSED branch before
  BANNED/SUSPENDED.
- D-21-B: notifyActiveMatches was a TODO stub → active Matches/conversations
  not closed on withdrawal (Gate ❌→✅); now finds active matches, sets
  UNMATCHED + closes conversation via chatService.closeConversationForMatch.
- D-21-C: cancelPendingActivityRegistrations was a TODO stub → future
  activity registrations not cancelled (Gate ❌→✅); now cancels pending/approved.
- D-21-D: anonymizeServiceHistory existed but was never wired into
  closeAccount → personal identity + counseling text leaked (Gate ❌→✅);
  wired in, anonymizes id_number/real_name with placeholder, plus
  counseling_followups content.
- D-21-E: closeAccount hard-threw on blockers → all cancelXxx unreachable
  (dead code contradiction); changed to informational log + continue cleanup.
- D-21-F: 7× assertTrue(true) fake-green tests → replaced with 11 real DB
  integration tests + 3 login-guard tests (Redis-decoupled via
  ReflectionTestUtils).

Cross-module defenses (RecommendationEngine/MatchEligibility/Capability
reject non-NORMAL) confirmed effective. Remaining (no silent scope creep):
matchmaker in-progress interception still hardcodes 0 (no table), data
retention delete flow is a product decision, privacy-settings endpoint stub.

Full regression: 650 tests / 0 failures / 0 errors / 3 skipped (baseline 648, +2).
 2026-09-18 10:12:34 +08:00
WorkBuddy 22c778b111 fix(analytics): close Phase 20 KEEP-VERIFY — 20 defects (fake metric definitions, hardcoded zeros, systemic soft-delete leak, module-wide 403, fake-green tests); add MetricCatalog + unified calculator + V043 grants + 51 guard tests; full regression 648 green 
口径造假：
- D1 实名人数用 user_profiles.profile_status != 'rejected' 冒充 → in_progress/need_edit 被算成已实名
- D3 Chat Started 直接数会话行数 → system/risk_notice 系统通知计入
- D4 Match 不分来源 → 活动 10 对互选被当 10 对成功引用（Gate）
- D6 Banner 曝光/点击硬编码 0 → 改返回 null + NOT_INSTRUMENTED 标记

死代码 / 占位：
- D5 活动漏斗全部硬编码 0 + 一行把"用户数"当"活动数"的死代码
- D20 getDailySummary(days) 无边界校验

系统性软删缺陷（D12）：
- BaseEntity.deletedAt 无 @TableLogic 且全局无逻辑删除配置
  → selectCount 把软删数据全部计入统计且无任何报错
- 修复 45 处统计调用显式 deleted_at IS NULL（2 轮复验，首轮漏扫 5 处次级路径：
  服务用户数分域明细 / 服务站统计 / 性别与年龄分布）

结构类：
- D15 Dashboard/用户漏斗/专项页各实现一遍同名指标且互不相同 → 抽 AnalyticsMetricsCalculator 统一口径

权限（Gate ⑩）：
- data_viewer 在 V001 定义却全仓从未授权（空壳角色）
- platform_admin 的 V015 扩展条件 id >= '20000000-…' 漏掉 V1.0 老权限段
  → 整个 Analytics 模块除 super_admin 外全员 403
- V043 授予两角色；刻意不授予 organization_admin（相关表无组织维度，无范围全量即越权）

假绿测试：
- 删除 AnalyticsServiceTest（5× assertTrue(true)，零断言）

新增：
- 指标字典 MetricCatalog（36 指标 × 12 字段）+ MetricStatus/MetricDefinition + docs/ANALYTICS_METRICS.md
- 守卫 51 用例：MetricCatalogGuardTest(11) / AnalyticsMetricsPhase20IntegrationTest(21)
  / AnalyticsSoftDeleteGuardTest(4，结构性) / AnalyticsPermissionGrantTest(6)
  / LocalEligibilityStatusVocabularyGuardTest(9)

跨阶段修复（缺陷本体属账号生命周期模块，按"发现即修"处理）：
- X1（Critical）注销写入 user_identities 状态域的 'failed' → 违反生产 V017 chk_elig_verify_status
  → 真实 PG 上 closeAccount 事务回滚、账号注销 100% 不可用（负向验证实测 DataIntegrityViolation）
- X2 checkCanWithdraw 资格拦截用 submitted/pending（该表永不可能出现）→ 死代码
- X3 测试库补 chk_elig_type/chk_elig_verify_status，暴露 ActivityPhase14Test 夹具违规

验证：
- 负向验证两轮共 7 处注入 → 7 个用例精准变红且无无关失败（一处被源码结构+运行时行为两层捕获）
- 还原后 md5sum 逐字节一致，NEG-VAL 残留 0
- 后端全量 648 tests / 0 failures / 0 errors / BUILD SUCCESS；admin-web vue-tsc EXIT=0
 2026-09-17 23:25:31 +08:00
WorkBuddy ff22c0d06a docs(phase19): cite the recovery tag instead of a moving commit sha 2026-09-17 22:25:10 +08:00
WorkBuddy d74f0ea099 docs(phase19): record tag/bundle location + commit sha in the recovery report 2026-09-17 22:24:56 +08:00
WorkBuddy 72b4e9b846 fix(ops-backend/service-station): close Phase 19 KEEP-VERIFY — 25 defects (unreachable announcements + station write privilege escalation; banner 500 + soft-delete leak; content publish dead-end; program approval dead-end/draft leak/zero counts/missing withdraw; zero-coverage test schema) ; add V042 announcement chain + 34 guard tests; full regression 602 green 
Backend
- Service Station: expose the 「临时公告」 chain end-to-end (V042 table/service existed
  but had ZERO endpoints -> both admin and miniapp could never reach it): 5 operator
  endpoints (list/create/update/publish/archive) + 1 user-facing read endpoint.
- Fix privilege escalation: create/update were guarded only by the class-level
  service_location.view, making V015's service_location.manage a never-used dead
  permission -> read-only roles could edit stations. Permissions moved to method level.
- Add ServiceLocationOperatorController to OperatorEndpointPermissionCoverageTest
  (it was NOT scanned, which is why the escalation stayed invisible).
- Add create/update validation (address NOT NULL + status CHECK only blew up on PG).
- Banner: validation failures threw bare RuntimeException -> 500; now VALIDATION_ERROR
  -> 400. Add deleted_at IS NULL to both home queries (soft-deleted banners stayed live).
- Content: operator list/detail split (admin «publish» was a dead end because it reused
  the published-only public API); category deactivate/reactivate; article body on edit.
- Service Program: operator member list now includes pending (approval reachable);
  active-only public visibility (draft leak); memberCount/pendingCount aggregation;
  withdraw/removeMember; reuse row on re-apply (UNIQUE(program_id,user_id) crash).
- MeController: add the missing GET /me/service-history route.

Schema / guards
- schema-h2.sql: add service_location_announcements (+ chk_announcement_status) and
  content_categories (+ uk_content_category_code, chk_category_status); move both into
  REQUIRED_IN_TEST_SCHEMA. MigrationValidationTest baseline 41 -> 42.
- V11PermissionMatrixContractTest now scans ALL migrations instead of hardcoding V015+V029.
- New: ServiceLocationAnnouncementPhase19IntegrationTest (24),
  ServiceLocationControllerPermissionContractTest (5),
  ServiceLocationAnnouncementConstraintParityTest (5, prod DDL <-> test schema <->
  backend whitelist <-> admin labels).

Frontend
- admin-web: ServiceStationView gains announcement management (statuses, create/edit,
  publish/archive with confirm, expiry); write buttons gated by service_location.manage.
- miniapp: U50 renders the 临时公告 block via new serviceStationService; page no longer
  talks to utils/api directly; announcement failure degrades to [] instead of erroring.
- ui-tests: U37 four-group + U50 announcement contracts; /home assertion retargeted to
  the service layer.

Verification
- Negative verification: 3 injected defects (drop soft-delete filter / drop test-schema
  CHECK / downgrade write endpoint to view) -> 4 named tests fail precisely, no
  collateral failures; restored, NEG-VAL residue = 0.
- Backend full regression: 602 tests, 0 failures, 0 errors, 3 skipped (baseline 420).
- admin-web: vue-tsc --noEmit EXIT=0, vite build EXIT=0. Miniapp: tsc --noEmit EXIT=0.
- ui-tests 14/14 PASS.
 2026-09-17 22:21:31 +08:00
WorkBuddy 3f60611a74 fix(appeal): close Phase 18 KEEP-VERIFY — 8 defects (zero-coverage appeals table + 6 fake-green tests; operator-notification FK crash; appealType CHECK; escalated dead-end; processAppeal race; soft-delete filters; hardcoded resolved; evidence contract); add AppealPhase18IntegrationTest (15); full regression 494 green 2026-09-17 21:07:25 +08:00
WorkBuddy 307c9efb36 fix(risk/report): close Phase 17 KEEP-VERIFY/EXTEND — 6 defects + 1 hidden compile defect (jakarta) + test-coverage blind zone (3 missing tables); add sensitive-access audit + ReportAndRiskPhase17IntegrationTest (7); full regression 488 green 2026-09-17 20:42:09 +08:00
WorkBuddy 556f95a562 fix(activity-mutual): close Phase 16 KEEP-VERIFY/EXTEND — 7 defects + 1 hidden compile defect (reopenEpisode) + 1 test-fidelity; unify dual-impl to confirmActivityInterest; INV-006 fix; +7 tests (481 green) 2026-09-17 16:47:10 +08:00
WorkBuddy 29112eed6e fix(activity): close Phase 15 KEEP-VERIFY/EXTEND — 2 critical + 14 other defects, 3 test-fidelity fixes 
Phase 15 (activity registration / review / checkin) verification closure.

Two CRITICAL defects, both "production always had it, test schema never did":
- D1: generateCheckinToken wrote the token to Redis with value "used", while
  scanCheckin rejects any token marked "used" -> every freshly issued check-in
  code was dead on arrival, staff scanning failed 100% of the time.
  Issuing now writes "issued"; "used" is set only after a successful scan.
- D9: UNIQUE(activity_id, user_id) exists in production, but re-registering
  after cancel/reject went through INSERT -> users who cancelled once could
  never register for that activity again (duplicate key 500). The H2 test
  schema lacked the UNIQUE, so the path had zero coverage. Now the original
  row is reused via an atomic conditional UPDATE.

Checkin (D2-D8):
- D2: duplicate check-in was read-then-write; now atomic checkinIfNotCheckedIn
  and repeat scans are idempotent (alreadyCheckedIn) without duplicate audit.
- D3: no data scope at all -> added activity staff assignment (V040) and
  assertCheckinScope; INV-020 (org relation grants nothing).
- D4: manualCheckin did not validate activity existence/cancellation.
- D5: getCheckinStats NPE on missing activity; remainingCount was semantically
  "free seats" -> replaced with notCheckedInCount.
- D6: malformed timestamp caused 500 -> 400. D7: constant-time signature compare.

Registration / review / cancel (D10-D13):
- D10: backend registration list only queried status='approved' -> the review
  screen never saw pending rows (review unusable).
- D11: registration_start_at / registration_end_at were never enforced.
- D12: cancel had no audit log (the only write without one).
- D13: reject/cancel were read-then-full-update -> rejectIfPending/cancelIfActive.

Capacity groups (D14, plan item 2): groups were stored and read back but never
participated in capacity -> incrementIfAvailable / decrementIfPositive now make
approved_count real; activities without groups are unaffected.

Frontend contract (F1-F2):
- F1: admin-web had no check-in surface at all -> ActivityCheckinView.vue
  (paste/scan token, manual check-in, stats) + /activities/checkin route & nav.
- F2: miniapp had no cancel-registration wrapper although the endpoint existed.

Test fidelity (T1-T3):
- T1: test schema gained UNIQUE(activity_id,user_id) + chk_reg_status +
  chk_checkin_status; two new guards parse the real V010 DDL and compare.
- T2: 3 assertTrue(true) placeholders in CheckinServiceTest (covering duplicate
  check-in, expiry, no-sensitive-plaintext) replaced with 5 real guards that
  parse the real source and real production DDL.
- T3: activity_registrations + activity_staff_assignments moved into
  REQUIRED_IN_TEST_SCHEMA; MigrationValidationTest baseline 39 -> 40.

Note: ActivityStaffAssignment must NOT extend BaseEntity - the table is a
composite PK (activity_id, operator_id); BaseEntity would add id/updated_at/
deleted_at to every INSERT and break on PostgreSQL (caught by
EntityMigrationParityTest immediately).

Tests: ActivityPhase15Test (21), CheckinServiceTest (5), 2 new constraint
guards. Negative verification: 3 injected defects -> 8 assertions turned red,
all restored.

Regression: 474 tests / 0 failures / 0 errors / 3 skipped / BUILD SUCCESS
(baseline 450). admin-web vue-tsc 0 error, vite build OK.
 2026-09-17 15:07:52 +08:00
WorkBuddy 755a570849 fix(activity): close Phase 14 KEEP-VERIFY/EXTEND — 11 backend + 3 frontend-contract + 2 test-fidelity defects 
Phase 14 (multi-type public-welfare activities) verification closure.

Backend (D1-D11):
- D1 CRITICAL: ActivityMapper.countRemainingCapacity had no @Select and no XML
  -> activity detail / register / approve all threw Invalid bound statement.
  Removed the shell; unified on ActivityRegistrationMapper.
- D2 CRITICAL: register & approve were read-then-write -> oversold under
  concurrency. Replaced with single atomic SQL (insertIfCapacityAvailable /
  approveIfCapacityAvailable, affectedRows==0 => full).
- D3 HIGH: activityType / orgRole never validated against production
  chk_activity_type / chk_org_role -> 500 only on PostgreSQL.
- D4 HIGH: ActivityListItemDTO had no activityType -> miniapp category tabs
  always empty (Gate failure).
- D5: male/female counts were fabricated as approvedCount/2; now real
  aggregation over user_profiles.gender.
- D6: minAge/maxAge was a TODO no-op; added assertAgeInRange.
- D7: registration_policy write-only; per-type eligibility now enforced
  (identity / local eligibility / gender quota).
- D8: POST /operator/activities dropped type, multi-org and capacity groups
  and set organizer_org_id to the operator user id; now delegates to
  createActivityFromAdmin (single implementation).
- D9: three operator write endpoints were mounted on the user-side
  ActivityController (dual implementation, zero callers) -> moved.
- D10: NPE on soft-deleted activity in getMyActivities; NPE/500 on missing
  admin form fields (Instant.parse(null)); tolerant time parsing added.
- D11: DTO lacked registeredCount / status / coverFileId / description
  -> card always showed 0 registered, "view detail", no cover, no summary.

Frontend contract (F1-F3):
- F1: miniapp category code "classroom" is not in the production whitelist
  (should be relationship_education); three types had no tab at all.
- F2: ui-activity-card read 7 fields that do not exist on the backend;
  now reads only real DTO fields, plus per-type badge and eligibility hint.
- F3: admin-web had no activity management page at all; added
  ActivityManagementView.vue (type-driven dynamic form, multi-org,
  capacity groups, publish/cancel) + /activities route and nav entry.

Test fidelity (T1-T2):
- T1: deleted two assertTrue(true) placeholder tests (testNoOverbooking,
  testNoDuplicateRegistration) that covered the exact broken invariants.
- T2: test schema lacked V023 columns/constraints and three tables; the
  activity tables were only in the OPTIONAL set (zero coverage) -> schema
  completed and moved to REQUIRED_IN_TEST_SCHEMA.

Tests: ActivityPhase14Test (23 cases), ActivityConstraintGuardTest (9 cases,
parses real db/migration/V*.sql plus miniapp source). Two negative-verification
rounds proved both guard families actually turn red.

Regression: 450 tests / 0 failures / 0 errors / 3 skipped / BUILD SUCCESS
(baseline 420). admin-web vue-tsc 0 error, vite build OK.
 2026-09-17 14:02:00 +08:00
WorkBuddy 7a2693b239 fix(counseling): close Phase 13A KEEP-VERIFY/EXTEND — 9 real defects + 4 test-infra defects 
Phase 13A（婚恋咨询与婚前辅导）复验关闭。历史 ✅ 不可信：counseling 模块存在双实现，
线上权威 CounselingService（167 行）零测试，唯一被测的 CounselingServiceApp（280 行）
是孤儿类（无注入点、无 HTTP 端点可达，生产永不执行）→ 已合并并删除。

生产缺陷（D1~D9）：
- D1 Critical 生产故障 · INV-013 泄露：咨询正文 userRequestSummary 原文被写入 WARN 日志
  （同 Phase 10 D2 一族）→ 改为只记 serviceId + riskLevel，绝不记正文。
- D2 High 越权：申请 userId 取自请求体而非认证主体 → Controller 强制 principal.id()。
- D3 High 并发：分配用「先读后写」→ 两管理员并发分配静默覆盖；
  改 assignIfPending 条件更新（WHERE status='pending' AND assigned_operator_id IS NULL），
  affectedRows=0 明确报错。
- D4 High 越权读：counseling_manager 经 getDetail 可读 followups 正文 → INV-013 绕过通道；
  manager 只加回 userRequestSummary，不装配 followups。
- D5 Medium：无状态机 → 「已完成」可被再取消；新增 ALLOWED_TRANSITIONS + assertTransition。
- D6 Medium 生产故障：GET /counseling/services/intro 后端不存在（且被 /services/{id} 吞掉）
  → 小程序 intro 页 100% 无数据；补端点并声明在变量路径之前。
- D7 Medium 契约：Admin 调用 5 条不存在的 /services/operator/*（404）+ 缺四态
  → 后端补 8 个运营端端点（@PreAuthorize），前端整体重写 + 注册 /counseling 路由。
- D8 Medium 契约：小程序字段名错（userRequestSummary/preferredServiceMode）+ 无幂等
  + my-counseling 未检查 success（403 静默空列表）+ 跳转不存在的页 + WXML 内调用方法。
- D9 Medium 测试保真度：测试库 counseling 两表零 CHECK 约束，
  counseling_services 未列入 EntityMigrationParityTest 任何集合（= 零覆盖）。

测试基建缺陷（T1~T4，不修即恒真假绿）：
- T1 守卫找不到 DDL：surefire workingDirectory = ${project.basedir}/.. 使 CWD 为仓库根，
  且 backend/ 是子目录非祖先 → 改为候选后缀 + 逐级向上定位，失败大声抛异常。
- T2 全限定 @com.baomidou...TableField(exist=false) 漏解析（假漂移警报）→ 正则放宽 + 回归守卫。
- T3 OperatorEndpointPermissionCoverageTest 扫到私有辅助方法（假阳性）→ 只扫描真实端点 + 自检。
- T4 两处测试自身逻辑缺陷：陈旧内存快照比较、Instant 纳秒 vs DB 微秒精度。

迁移：新增 V039（counseling 幂等键 + 三个 CHECK 白名单显式固化），基线 38 → 39。

负向验证两轮（均已还原，grep -c NEG-VAL = 0）：
- ① DDL 常量守卫：从 V039 移除 'scheduled' → 常量守卫精准变红
  （证明解析的是真实 DDL 而非副本）。
- ② 行为守卫：注入「并发回退先读后写 / manager 读 followups / 状态机守卫失效」
  → 三个用例各自精准变红。

测试：新增 CounselingPhase13ATest（25）+ CounselingConstraintGuardTest（5，解析真实 DDL）；
删除孤儿类及其 3 个 DTO、删除测死代码的 CounselingDataScopeTest。
全量回归：Tests run: 420, Failures: 0, Errors: 0, Skipped: 3 / BUILD SUCCESS（4:49）。
基线 391 → 420（净 +29）。admin-web vue-tsc EXIT=0；miniapp tsc EXIT=0。

Tag: phase13a-recovery-2026-09-17
Report: docs/Phase-13A-V11-Recovery-Completion-Report-2026-09-17.md
 2026-09-17 01:25:12 +08:00
WorkBuddy 5b3bad326e fix(matchmaker,referral): close Phase 13 KEEP-VERIFY/EXTEND — 7 real defects 
KEEP-VERIFY/EXTEND 复验（历史 ✅ 不可信；测试覆盖函数 ≠ 覆盖运行时路径）。
逐条核对运行时路径与生产约束后修复 7 处缺陷，其中 3 处「只在真实 PostgreSQL 上炸」。

D1 (Critical, 生产故障)
  matchmaker_recommendations.status 代码写 liked/passed/confirmed_mutual，
  而 V009 chk_rec_status 只允许 pending_view/viewed/responded 且从未扩展
  → 真实 PG 上用户接受/跳过推荐必崩。测试库未复刻该 CHECK，故长期静默。

D2 (Critical, 生产故障)
  控制器缺省 serviceType="recommendation" 不在 chk_matchmaker_service_type
  白名单 → 提交红娘服务必崩。
  D1+D2 同修于新增 V038__align_matchmaker_status_constraints.sql
  （沿用 V018/V020/V025 的 DROP CONSTRAINT IF EXISTS + 重加白名单惯例）。

D3 (High, 越权写)
  MatchmakerServiceService.recommendPartner 无 assignedOperatorId 归属校验，
  且可把用户推荐给自己 → 补校验 + 禁止自我推荐。

D4 (High, 越权读)
  getServiceDetail 无归属校验且返回完整用户资料（userId/description/
  userProfile/followups）→ 签名加 requesterId，按调用者返回完整/脱敏视图。

D5 (High, 越权写)
  ReferralService 的 userId 取自请求体且零授权校验 → 新增
  assertMatchmakerServesUser，在 referToServiceStation/referToActivity 入口
  强制校验存在进行中的服务关系。

D6 (High, 生产故障)
  MatchmakerFollowupMapper.findByServiceId 只有声明无 @Select
  → Invalid bound statement (not found)，getServiceDetail 必崩。
  该缺陷由本轮新增测试首次暴露（历史测试从未调用该路径）→ 补 @Select。

D7 (Medium, 契约缺口)
  admin-web 5 处 API 缺口（4 条 404 + C 端端点被运营端误用）+ 小程序缺转介服务层
  → 后端补 GET /operator/matchmaker/recommendations 与
  GET /referrals/operator/history 两个运营端端点；前端改路径、
  assign 改 POST + body operatorId；新增 miniapp/services/referralService.js。

测试保真度
  - 测试库补 matchmaker_services/recommendations/followups 三表（此前该模块零真实覆盖），
    约束与 V038 对齐；EntityMigrationParityTest 三表移入 REQUIRED_IN_TEST_SCHEMA；
    MigrationValidationTest LATEST_MIGRATION 37 → 38。
  - 新增 MatchmakerReferralPhase13Test（17 用例，@SpringBootTest 走真实方法路径 + 真实落库）。
  - 重写 MatchmakerServiceTest（6 用例）：直接解析 db/migration/V*.sql，按文件名顺序应用
    DROP CONSTRAINT IF EXISTS + ADD CONSTRAINT 取"最终有效"白名单，并内置解析器自检。
    【关键】首版守卫硬编码白名单副本，注入 D1 后仍全绿（对漂移零反应）→ 必须解析真实 DDL。
  - 删除 3 处假测试占位（MatchmakerServiceTest.testMatchmakerCannotLike 的 assertTrue(true)；
    CoreInvariantsTest INV-007 断言的是状态常量、INV-010 亦为 assertTrue(true)），
    已如实改名并注明行为验证位置。

负向验证（两轮）
  注入 D1（回退 chk_rec_status 白名单）→ 行为守卫 DataIntegrityViolation
  + 常量守卫 Multiple Failures (3 failures) → Tests run: 23, Failures: 1, Errors: 2
  / BUILD FAILURE，守卫有效性得证。缺陷已还原。

全量回归（离线容器，实测）
  [WARNING] Tests run: 391, Failures: 0, Errors: 0, Skipped: 3
  [INFO] BUILD SUCCESS / Total time: 04:59 min
  Phase 12 基线 371 → 391（净 +20）。

报告：docs/Phase-13-V11-Recovery-Completion-Report-2026-09-16.md
Phase 13A 未自动进入，等待用户确认。
 2026-09-17 00:01:53 +08:00
WorkBuddy c607e87b45 fix(relationship): Phase 12 复验 —— 修复实体↔生产迁移系统性漂移（生产故障族）/ 双路径 episode 与 NPE / 测试库缺表致首页零覆盖 / 计划项 2 未实现 
KEEP-VERIFY 复验发现并修复 4 处真实缺陷：
- D1（Critical，生产故障·同一族系统性缺陷）实体与 db/migration 系统性漂移：23 表 50 列在生产
  迁移中不存在，而测试库 schema-h2.sql 手写补齐 → 全量绿=假绿。其中 11 表属"INSERT 必崩"级
  （缺 id/created_at/updated_at，被 AutoFillHandler 强制填充），含 matches.episode、user_preferences
  权重列、permissions/roles、consent_records、risk_events 等；真实 PG 上 Match 创建、择偶偏好保存、
  权限/角色写入、同意记录、风险事件全部失败。修复：新增 V037（幂等 ADD COLUMN IF NOT EXISTS
  + 存量 episode 回填）。
- D2（High）RelationshipService.createMatch 不写 episode/lastActivityAt，与 MatchService 双路径
  填充不一致；且 episode 为 null 时 m.getEpisode()+1 对 Integer 自动拆箱抛 NPE。修复：统一按
  AGENTS.md #20 推导 + null 兜底。
- D3（High）测试库缺 user_identities 等 7 表 → GET /api/v1/home 在测试环境 BadSqlGrammar，
  首页接口零回归覆盖。修复：schema-h2.sql 补齐 7 表。
- D4（Medium）Phase 12 计划项 2 实际未实现：getting_to_know 分支只产出 counseling。修复：
  HomeApplicationService 补 premarital 引导（与 counseling 并列）。

测试保真度：
- 新增永久守卫 EntityMigrationParityTest（5 用例）：把"实体字段 ⊆ 生产迁移累积列"与"测试库覆盖
  关键业务表"固化为自动不变量，区分"必崩级/潜伏级"；经两轮负向验证证明会变红（第一轮分级过窄
  已修正：episode 由"潜伏级"纠正为"必崩级"）。
- 新增 RelationshipProgressPhase12Test（14 用例，含 8 线程并发确认）与
  HomePremaritalGuidancePhase12Test（5 用例）。
- 删除 2 个假测试类中的 assertTrue(true) 占位（AGENTS.md 明令禁止），改走真实服务方法验证。

Gate：单方确认不改变最终关系 ✅ / 双确认后才 successful ✅ / 历史 Episode 可追踪 ✅ /
新 Episode 不恢复旧聊天权限（conversations.match_id UNIQUE）✅ / 非成员不可代确认 ✅

全量回归：371 tests / 0 failures / 0 errors / 3 skipped / BUILD SUCCESS（71 类，5m07s）
 2026-09-16 23:01:13 +08:00
WorkBuddy 52e11a6c65 fix(notification): Phase 11 复验 —— 修复通知类型违反生产 CHECK 约束（生产故障）/ 幂等无库兜底 / 婚前辅导通知未接线 
复验模式：KEEP-VERIFY/EXTEND（不轻信历史 ✅，逐条核对运行时路径与生产约束）。

== D3（Critical · 生产故障）通知 type 全量违反生产 CHECK 约束 ==
NotificationService 的 10 个业务模板传入复合 type
（appointment_submitted / counseling_assigned / le_verified / referral_created 等），
而生产 V020__extend_notification_types.sql 的 chk_notification_type 白名单只允许 12 个短名。
真实 PostgreSQL 下这些通知全部落库失败；测试库 schema-h2.sql 未复刻该约束，故历史 334 测试全绿 → 假绿。
修复：
- type 归一为白名单短名，业务细分交由 businessType/title 承担：
  appointment_submitted|_confirmed|_changed -> appointment
  counseling_assigned|_scheduled|_completed  -> counseling
  premarital_scheduled                       -> premarital
  le_need_evidence|le_verified|le_rejected   -> local_eligibility
  referral_created                           -> matchmaker
- 重写 buildPrivacySafeCopy：原以复合 type 为 key 的 Map.ofEntries 归一后产生重复 key 会抛
  IllegalArgumentException（运行期崩溃），改为按短名 switch。
- schema-h2.sql 补齐生产 CHECK 约束（测试保真度下沉）。

== D1（High）通知幂等仅 SELECT-then-INSERT，无 DB 兜底 ==
NotificationMapper.findExisting 为查后插，并发下可重复落库。
修复：新增 V036__add_notification_idempotency_index.sql，建 partial unique index
uk_notification_business(user_id, type, business_type, business_id, content)
WHERE business_id IS NOT NULL（先按 created_at ASC 去重存量数据）。
schema-h2.sql 同步唯一索引（H2 无 partial index，标准 UNIQUE 对 NULL 多行共存语义等价）。

== D2（Medium）婚前辅导通知未接线 ==
premaritalScheduled 零调用；CounselingServiceApp.scheduleCounseling（P13A-T22 管理端路径）完全不发通知。
修复：CounselingService.schedule 与 CounselingServiceApp.scheduleCounseling 均按 service_type 分流，
premarital 走 premaritalScheduled，其余走 counselingScheduled；
管理端路径包 try/catch —— 通知失败不回滚排期事实。

== 测试（基线 334 → 346，+12） ==
- NotificationPhase11IntegrationTest（5）：回滚不发通知（Gate 前半）、提交后可见、
  顺序重试幂等、绕过 findExisting 由唯一索引兜底抛异常且最终仅 1 条、business_id IS NULL 多行共存。
- NotificationPrivacyCopyTest（5）：反射校验 4 类隐私文案，咨询类不含「咨询」且标题为「服务通知」，
  不泄露姓名/时间/证件类型；未知 type 安全回退；ALLOWED_TYPES 守卫。
- NotificationTemplateTypeGuardTest（2）：关键 —— 调用模板方法本身（而非直接传字面量 send），
  断言 10 个模板全部以白名单短名落库。
- MigrationValidationTest：LATEST_MIGRATION 35 -> 36，DDL 校验补 CREATE UNIQUE INDEX 识别。

== 负向验证（两轮，含第一轮失效教训） ==
第一轮：注入复合 type 后 NotificationPhase11IntegrationTest 仍全绿 ——
  因该测试直接 send(..., "appointment", ...) 传字面量，不经过模板方法，属无效守卫。
第二轮：改为 NotificationTemplateTypeGuardTest 调用模板本身，注入后精准抛
  DataIntegrityViolation，守卫有效性得证。教训已写入报告：守卫必须经负向验证证明会变红。

== 证据 ==
全量回归 346 tests / 0 failures / 0 errors / 3 skipped / BUILD SUCCESS
（离线 Docker：danyang-match-backend-test:latest + 卷 danyang-m2-cache）
迁移校验 12/12；定向 19/19。

== Gate 判定 ==
回滚事务不发通知 ✅ ｜ 通知失败不回滚核心事实 ✅ ｜ 隐私安全文案 INV-010 ✅
已提交事务通知失败可重试且不产生重复业务记录 ✅（修复后成立）

== 遗留 gap（登记不自动实现） ==
sendExternal 在事务内同步执行外部副作用，当前为占位实现；接入真实微信推送前必须改为
@TransactionalEventListener(AFTER_COMMIT) / REQUIRES_NEW —— 全仓当前无任何 afterCommit 机制。
 2026-09-16 21:50:56 +08:00
WorkBuddy fd7195ecb0 fix(safety): Phase 10 复验 —— 修复图片上传魔数校验死代码（生产故障）/ 高风险词写入日志 
对 Phase 10（Image & Chat Safety）执行 KEEP-VERIFY 复验：不采信历史 ✅，逐条核对
图片私有访问、类型/大小/魔数校验、审核与水印、风险识别/举报/违规处理、日志脱敏。
Gate 判定通过，发现并修复 2 处缺陷，并推翻旧报告 2 项安全声明。

D1（Critical，生产故障）图片上传的类型与魔数校验整条是死代码
  - PhotoService.uploadAndSave 用 `header = new byte[Math.min(12, stream.available())]`
    仅分配零值缓冲，从不调用 stream.read(...) → validateMagicNumber 对任何文件都收到
    12 个 0x00 → 恒返回 false → 恒抛"无效的图片文件头"。
  - 影响链（已确认线上触发）：小程序 local-eligibility-supplement.ts 以
    photoType='local_eligibility' 调 POST /photos/upload → 本地资格佐证上传 100% 失败。
  - 次生问题：isAllowedMimeType（image-only 白名单）在 local_eligibility 分支之后
    无条件执行，即使魔数修好，application/pdf 仍被挡，而 FeatureFlagConfig
    的 localEvidenceAllowedTypes 明确允许 pdf；另有未 mark 的 stream.reset() 无效调用。
  - 测试为何没发现：PhotoServiceTest 只对 validateMagicNumber 传手写字节做单测，
    从不经过 uploadAndSave，该路径缺陷完全无覆盖。
  - 修复：改用 try(InputStream){ header = s.readNBytes(12); } 真正读取；
    MIME 校验按 photoType 分流（local_eligibility 走自己的白名单含 pdf）；
    魔数校验仅对图片执行、PDF 跳过；移除死 reset()；上传用全新流与校验流隔离；
    generateFileKey 重载按 Content-Type 推断扩展名，PDF 不再拿到误导性 .jpg key。

D2（Low）高风险词写入普通日志
  - ChatSafetyService.scanContent 把命中的高风险词原样写入 log.warn，
    与同文件 CONTACT 分支"只记 pattern 不记内容"矛盾，且违反本阶段 Gate。
  - 修复：改为 log.warn("Chat content matched HIGH_RISK keyword (blocked)")。

推翻的旧声明（docs/Phase-10-Completion-Report.md）
  - 「伪装恶意图片被拒（Magic Number校验）」不成立：该校验运行时恒失败，
    连合法图片也一律被拒（fail-closed 到不可用），安全属性并非由它提供。
  - 「EXIF清理预留（PhotoService.stripExif占位）」不存在：全仓库无该符号。
    且 getUploadToken 客户端直传路径服务端无法介入清理 EXIF。
  - 已在旧报告顶部加状态注记，指向本次复验报告。

测试
  - 新增 PhotoUploadValidationTest（6 用例），用 MockMultipartFile 真正走一遍
    uploadAndSave：合法 JPEG/PNG 通过且确实调用 putObject、伪造 Content-Type 被魔数挡、
    PDF 佐证不被 image-only 挡且 key 以 .pdf 结尾、非法 MIME 被拒、超大被拒。
  - 负向验证：临时把魔数守卫改为 if(false && ...) → 恰好 spoofedContentTypeRejectedByMagic
    一个用例变红，其余 5 个仍绿；随后还原并 grep 确认无 NEGATIVE-CHECK-TEMP 残留。
  - 全量回归（离线容器 danyang-match-backend-test，mvn -o test）：
    334 tests / 0 failures / 0 errors / 3 skipped / BUILD SUCCESS（基线 328，+6）。

登记 gap（不自动实现，待产品确认）
  - 图片水印：全仓库无实现；主计划未列硬需求，仅"开发执行计划修订版"提"高级动态水印"。
  - EXIF 清理：旧声明不存在；客户端直传路径服务端无法介入，需改为接收侧重编码。
  - 直传凭证路径（getUploadToken）无服务端内容校验且不落 DB 记录 → 后续 getSignedUrl
    归属校验会查不到 UserPhoto 行，属契约层缺口，超出本阶段安全 Gate 范围。
  - 违规自动升级：ChatSafetyService 注释明确属 Phase 17 范围，当前 high 仅拦截不落库。

文档：新增 Phase-10-V11-Recovery-Completion-Report-2026-09-16.md；
旧报告加状态注记；主计划 Phase 10 复选框置 ✅；STATUS / JIRA_BACKLOG 同步（334 tests）。

Phase 11 未启动，按计划每阶段出报告后停止等待确认。
 2026-09-16 21:00:52 +08:00
WorkBuddy 03a03c7c11 fix(chat): Phase 09 复验 —— 修复 WebSocket 下行死链路 / Unmatch 后可发消息 / Block 未落库 / 聊天 IDOR 
对 Phase 09（Text Chat）执行 KEEP-VERIFY/EXTEND 复验：不采信历史 ✅，逐条核对
REST 发送 + WebSocket 接收、幂等、重连、未读计数、非 active Match 禁聊、Unmatch/
Block 即时禁聊、活动双确认会话复用。共发现并修复 4 处真实缺陷 + 1 处规范不符。

D1（Critical）WebSocket 下行整条链路是死代码
  - 无 HandshakeInterceptor，无任一环节把 userId 写入 session attributes，
    extractUserId 恒返回 null；sendToUser/sendToConversation 全仓库零调用点。
  - 新增 ChatHandshakeInterceptor：从 ?token= 或 Authorization: Bearer 解析 JWT，
    校验通过后写 session attr userId；失败 fail-closed 拒绝握手；仅接受 type=access。
  - 重写 ChatWebSocketHandler：按规范 {event,data} 信封推送；未认证连接直接
    close(NOT_ACCEPTABLE)；同账号重连由新连接接管、旧连接 close(NORMAL)；
    上行仅支持 ping->pong，禁止客户端自订阅。
  - ChatService.sendText 落库+递增未读后推送 message.new；RiskNotice 同推。
  - 移除 WebSocketConfig 的 .withSockJS()（SockJS 改写端点致原生 WS 无法直连）。

D2（Critical）Unmatch 后仍可发消息
  - unmatch 只改 Match，不联动会话；sendText 仅校验 conv.status==active。
  - 新增 ChatService.assertMatchAlive(conv)：按 matchId 查 Match，状态非
    {active,getting_to_know} 抛 "匹配已解除，无法发送消息"。sendText 在 conv.status
    校验之后调用。
  - MatchController.unmatch 落库后调 chatService.syncFromMatch 推 conversation.closed。
  - syncFromMatch 重写：alive 恢复 active；非 alive 只推 conversation.closed，
    不改 status（会话仍以灰显态留在列表，属未授权产品决策，登记待决）。

D3（Critical）RelationService.block 调静态 MatchService.unmatch 只改内存态不落库
  - 结果：Block 后 matches.status 仍为 active。
  - 补 matchMapper.updateById(m) 持久化 + chatService.closeConversationForMatch(
    CONVERSATION_STATUS_BLOCKED) 通知对端。RelationshipService.block 同步补会话关闭通知。

D4（Critical 安全）聊天 4 个接口无成员校验 -> IDOR 越权读私聊
  - MessageController.load/markRead、ConversationController.detail/unread 补
    chatService.isMember(conversationId, userId) 校验；非成员返回
    ApiResponse.notFound()（body 级 success=false code=NOT_FOUND，HTTP 200，
    与 MatchController.authorizedMatch 同约定），不区分"不存在"与"无权"以防探测。

D5（Medium）上行订阅/信封不符规范
  - REST API 接口详细设计规范.md 第124-126章要求禁止客户端自订阅，事件名
    message.new/message.read/conversation.closed，信封 {event,data}。已对齐。

测试保真度 + 测试补齐
  - schema-h2.sql 新增 uk_message_client_id(sender_id, client_message_id)，
    对齐生产 V004:71；H2 不支持过滤索引故省略 WHERE（NULL 语义等价）。
    否则 service 的 SELECT-then-INSERT 竞态在测试中无 DB 兜底，形成测试盲区。
  - 新增 ChatPhase09IntegrationTest（13 用例）：幂等重试不重复投递、推送信封字段、
    未读增减清零、Match unmatched 拒发、Unmatch/Block 端点即时禁聊+收 closed 事件、
    活动双确认复用同一消息体系、未形成 Match 无法建会话、非成员读消息/详情/未读
    被拒且不泄露正文、DB 唯一索引兜底重复 clientMessageId、无幂等键系统消息可共存。
  - 新增 ChatWebSocketHandlerTest（7 用例）：userId 路由、信封格式、未认证关连、
    断开不推、重连替换旧连接、上行仅 ping。

验证证据
  - 全量回归（离线容器 danyang-match-backend-test，mvn -o test）：
    328 tests / 0 failures / 0 errors / 3 skipped / BUILD SUCCESS（基线 308，+20）。
  - 负向验证：临时注释 assertMatchAlive + 两处 isMember，恰好 3 个用例变红，
    证明守卫有效；随后还原并 grep 确认无残留。

文档
  - 新增 Phase-09-V11-Recovery-Completion-Report-2026-09-16.md（完整报告）。
  - Phase-09-V11-Completion-Report.md 顶部加状态注记，推翻其 4 条 WebSocket 结论。
  - 计划/STATUS/JIRA_BACKLOG 同步 Phase 09 完成态与测试总数（328）。

Phase 10 未启动，按计划每阶段出报告后停止等待确认。
 2026-09-16 16:42:48 +08:00
WorkBuddy 4e478ce6ba fix(invariants): Phase 08 复验 —— 修复 INV-004 搜索/详情绕过 Block + matches.source 取值越界 
复验（KEEP-VERIFY）在既有 Phase 08 实现中发现 3 处真实缺陷并修复：

D1（严重）DiscoverService.searchUsers 不含查看者身份，查询条件既无排除自己也无任何 Block 条件，被拉黑者仍可被搜到并点开 —— INV-004「搜索不可见」整体失效。改为接收 viewerUserId，排除自己，并用新增的 UserRelationMapper.findBlockedUserIdsBothWays（UNION 双向）一次取回后 notIn 批量排除，避免 N+1。

D2 详情路径（DiscoverService / UserProfileService）均无 Block 判断，知道 userId 即可读到被拉黑者资料。两条路径统一加双向判断并返回 null，由 Controller 转 404（不用「已隐藏」以免被用来探测存在）；UserProfileService 中该判断排在隐私设置之前，拉黑不能被对方「允许搜索可见」覆盖。

D3（潜伏，仅 PostgreSQL 暴露）RelationshipService.like() 把关系类型值 "liked" 当 match source 写入 matches.source，违反 V003:91 的 chk_match_source。测试库 schema-h2.sql 该列是裸 VARCHAR(32)、无约束，故长期静默通过。改传 MATCH_SOURCE_USER_LIKE。canonical 路径 RelationService.like 传值本就正确，线上未触发。

测试保真度对齐：schema-h2.sql 补 chk_match_source（与 V003 逐字对齐）与 user_profiles 表（此前测试库无此表，资料查询无法端到端测试）。

新增 DiscoverBlockInvisibilityIntegrationTest（9 用例，@SpringBootTest 真实 H2 + 真实 SQL），覆盖正向/反向拉黑、对称性、排除自己、详情 404 与查自己不受限。P0FixVerificationTest 补 source 回归守卫。DiscoverServiceTest 同步构造签名。

负向验证：临时还原 D3 修复后，4 个用例因 CHK_MATCH_SOURCE 约束违反报错，证明守卫有效。

证据：后端全量 308 tests / 0 failures / 0 errors / 3 skipped（基线 299）；UI 契约 13/13 GREEN。报告 docs/Phase-08-V11-Recovery-Completion-Report-2026-09-16.md。

同步 STATUS.md / JIRA_BACKLOG.md（EPIC-P08-V11 置 ✅）/ 客户需求基线计划 Phase 08 复选框；首轮报告加状态注记（其「Known Issues 无」已被推翻）。.gitignore 忽略 .workbuddy/。
 2026-09-16 15:11:38 +08:00