@geoly-ai/social-hub-cli 0.2.2 → 0.2.4

package/CHANGELOG.md

@@ -4,6 +4,43 @@ All notable changes to `@geoly-ai/social-hub-cli` are documented in this file.
 
 Release workflow: `pnpm release:cli` → `pnpm publish:cli`.
 
+## [0.2.4] - 2026-07-16
+
+### Added
+
+- 帖子审核五层交付——SDK postReview + CLI content post-review + skill
+- 需求A P1 收尾——behavioralHealth advisory feature flag(默认关)
+- 需求A P1——behavioralHealth 客户端方法 + query 子命令
+- 互动时序 P2——批量调度上下文 query + correlationKeys + §13 收口
+- 互动时序去同质化 Hub P1——canonical 事件/时区溯源/活动聚合
+
+## [0.2.3] - 2026-07-14
+
+content review 命令(评论 LLM 复审+优化,层③);reddit list 排序参数;brands system-update;seqNo 字段
+
+### Added
+
+- 评论 LLM 复审+优化服务——发布前三层 gate 的层③(独立模型对抗式复审)
+- Reddit 指标快照剔除内部自建品牌 + brands.is_internal 产品化
+- 帖子表现新增数字序列号 seqNo——可编辑/列表展示/可排序
+- 草稿一稿一投——entry 成功自动 published,禁止再排期(含并发安全与历史回填)
+- publishing plan 状态改为 entries 聚合派生 + status 过滤真正生效
+- CLI/API accounts reassign 命令(账号跨团队迁移产品化)
+- 草稿变更追踪 + 乐观并发(updatedAt/version,防丢更新)
+- 账号注册时间 registeredAt(可编辑)+ 派生只读 age
+- 通用抓取加 scrollCount(无限滚动页懒加载)
+
+### Fixed
+
+- 收官清扫——cron addBulk/seed 契约/快照幂等/intel partial 终态对齐
+- sanction 解封语义收紧——枚举化+全写入口归一,web"解除"按钮从未真正解封
+- 终态一致性——complete/fail/skip 幂等 + fail/skip 事务化 + cron 任何终态都续期
+- Firecrawl 兜底改用低产量重试(自托管不支持 actions),去掉 scrape scrollCount
+
+### Documentation
+
+- 说明 REVIEW_LLM_BASE_URL 三 provider 通用+allowlist 要求
+
 ## [0.2.2] - 2026-07-03
 
 skills install 未指定 agent 时默认自动探测机器上已有的 agent 目录安装(尊重 --global/--project;TUI 空选同样回退);不再报错。

package/dist/cmd-manifest.json

@@ -1,6 +1,6 @@
 {
-  "generatedAt": "2026-07-03T08:48:48.566Z",
-  "cliVersion": "0.2.2",
+  "generatedAt": "2026-07-16T10:14:09.932Z",
+  "cliVersion": "0.2.4",
   "commands": [
     "accounts",
     "accounts brand-bindings",
@@ -25,6 +25,7 @@
     "accounts personas",
     "accounts personas create",
     "accounts personas list",
+    "accounts reassign",
     "accounts status-snapshots",
     "accounts status-snapshots create",
     "accounts status-snapshots list",
@@ -71,6 +72,7 @@
     "brands list",
     "brands system-create",
     "brands system-list",
+    "brands system-update",
     "brands update",
     "calendar",
     "calendar list",
@@ -102,6 +104,9 @@
     "config init",
     "config set",
     "config show",
+    "content",
+    "content post-review",
+    "content review",
     "context",
     "context account",
     "context current",
@@ -248,6 +253,8 @@
     "query",
     "query account-pool-context",
     "query account-pools",
+    "query interaction-behavior-health",
+    "query interaction-schedule-context",
     "query subreddit-candidates",
     "reddit",
     "reddit batch-upsert",

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":";AACA,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AA2BpC,QAAA,MAAM,OAAO,SAAgB,CAAC;AAw9C9B,8DAA8D;AAC9D,wBAAgB,4BAA4B,CAC1C,IAAI,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,MAAM,GAAG,IAAI,CAOf;AAk2ED,OAAO,EAAE,OAAO,EAAE,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":";AACA,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AA2BpC,QAAA,MAAM,OAAO,SAAgB,CAAC;AAgiD9B,8DAA8D;AAC9D,wBAAgB,4BAA4B,CAC1C,IAAI,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC5B,MAAM,GAAG,IAAI,CAOf;AAy3ED,OAAO,EAAE,OAAO,EAAE,CAAC"}

package/dist/index.js

@@ -334,6 +334,8 @@ reddit
     .option("--repost-type <t>", "repostType filter (none|first|second)")
     .option("--contains-brand-keyword <v>", "true | false | unset")
     .option("--primary-product <text>", "primary product contains match")
+    .option("--sort-by <f>", "score|comments_count|posted_at|view_count|seq_no")
+    .option("--sort-dir <d>", "asc|desc")
     .option("-n, --limit <n>", "limit per page", "100")
     .option("--offset <n>", "pagination offset")
     .option("--all", "fetch all pages (200 per request until total reached)")
@@ -351,6 +353,8 @@ reddit
         repostType: opts.repostType,
         containsBrandKeyword: opts.containsBrandKeyword,
         primaryProduct: opts.primaryProduct,
+        sortBy: opts.sortBy,
+        sortDir: opts.sortDir,
         ...redditAggregateQueryParams(opts),
     }, { fetchAll: Boolean(opts.all) });
     console.log(JSON.stringify(res, null, 2));
@@ -449,6 +453,43 @@ jobs
     console.log(JSON.stringify(res, null, 2));
 });
 // --- drafts & reports ---
+const content = program
+    .command("content")
+    .description("内容质量(发布前 LLM 复审)");
+content
+    .command("review")
+    .description("POST /content-review — 评论发布前独立 LLM 复审+优化(层③);degraded 时降级本地自审")
+    .requiredOption("-t, --team <teamId>", "Team UUID")
+    .requiredOption("-j, --json <json>", "{ accountId, draft, commentType, thread:{subreddit,postTitle,...}, rewrite?, stage?, attempt? }")
+    .action(async (opts) => {
+    let body;
+    try {
+        body = JSON.parse(opts.json);
+    }
+    catch {
+        console.error("Invalid -j / --json");
+        process.exit(1);
+    }
+    const res = await requireClient().contentReview(opts.team, body);
+    console.log(JSON.stringify(res, null, 2));
+});
+content
+    .command("post-review")
+    .description("POST /post-review — 帖子(title+body)发布前独立 LLM 审核(七维,独立模型 key);仅审帖子正文,评论审核用 content review")
+    .requiredOption("-t, --team <teamId>", "Team UUID")
+    .requiredOption("-j, --json <json>", "{ accountId, title, body, subreddit, subredditRules?, rewrite?, stage?, attempt? }")
+    .action(async (opts) => {
+    let body;
+    try {
+        body = JSON.parse(opts.json);
+    }
+    catch {
+        console.error("Invalid -j / --json");
+        process.exit(1);
+    }
+    const res = await requireClient().postReview(opts.team, body);
+    console.log(JSON.stringify(res, null, 2));
+});
 const drafts = program.command("drafts").description("内容草稿");
 drafts
     .command("list")
@@ -555,7 +596,7 @@ brands
 brands
     .command("system-create")
     .description("POST /system/brands (preferred)")
-    .requiredOption("-j, --json <json>", "{ name, slug? }")
+    .requiredOption("-j, --json <json>", "{ name, slug?, isInternal? }")
     .action(async (opts) => {
     let body;
     try {
@@ -568,9 +609,27 @@ brands
     const res = await requireClient().createSystemBrand({
         name: String(body.name ?? ""),
         slug: body.slug,
+        isInternal: body.isInternal,
     });
     console.log(JSON.stringify(res, null, 2));
 });
+brands
+    .command("system-update")
+    .description("PATCH /system/brands/:id")
+    .requiredOption("--brand <uuid>", "Brand UUID")
+    .requiredOption("-j, --json <json>", "{ name?, status?, isInternal? }")
+    .action(async (opts) => {
+    let body;
+    try {
+        body = JSON.parse(opts.json);
+    }
+    catch {
+        console.error("Invalid -j / --json");
+        process.exit(1);
+    }
+    const res = await requireClient().updateSystemBrand(opts.brand, body);
+    console.log(JSON.stringify(res, null, 2));
+});
 brands
     .command("list")
     .description("GET /teams/:teamId/brands (compat)")
@@ -821,10 +880,12 @@ plans
     .requiredOption("-t, --team <teamId>", "Team UUID")
     .option("--brand <uuid>", "brandId 过滤（推荐）")
     .option("--campaign <uuid>", "campaignId 过滤（legacy）")
+    .option("--status <status>", "派生状态过滤（scheduled/running/succeeded/failed/cancelled;无 entries 的计划保留存储值 draft/exported）")
     .action(async (opts) => {
     const res = await requireClient().listPublishingPlans(opts.team, {
         brandId: opts.brand,
         campaignId: opts.campaign,
+        status: opts.status,
     });
     console.log(JSON.stringify(res, null, 2));
 });
@@ -1253,7 +1314,7 @@ accounts
     .description("PATCH /accounts/:id — 更新 handle/status/karma/accountAgeDays 等核心字段。安全时段/时区/风险级别请用 `compliance risk-put`;浏览器环境请用 `browser-envs`")
     .requiredOption("-t, --team <teamId>", "Team UUID")
     .requiredOption("--account <uuid>", "Account UUID")
-    .requiredOption("-j, --json <json>", "{ handle?, profileUrl?, status?, externalId?, externalRef?, personaId?, karma?, accountAgeDays?, workflowStage?, metricsSource?, metricsUpdatedAt?, credentialsRef?, primaryProducts? }")
+    .requiredOption("-j, --json <json>", "{ handle?, profileUrl?, status?, externalId?, externalRef?, personaId?, karma?, accountAgeDays?, registeredAt?, workflowStage?, metricsSource?, metricsUpdatedAt?, credentialsRef?, primaryProducts? }")
     .action(async (opts) => {
     const body = JSON.parse(opts.json);
     const misplaced = detectMisplacedAccountFields(body);
@@ -1264,6 +1325,18 @@ accounts
     const res = await requireClient().updateAccount(opts.team, opts.account, body);
     console.log(JSON.stringify(res, null, 2));
 });
+accounts
+    .command("reassign")
+    .description("POST /accounts/:id/reassign — 把账号从 -t <team> 移到 --to <team>。运营配置(浏览器环境/人设/子版亲和/封禁)跟账号走,纯历史留原队,原队待跑任务取消。目标队撞唯一键返回 409")
+    .requiredOption("-t, --team <teamId>", "源 Team UUID(账号当前所属)")
+    .requiredOption("--account <uuid>", "Account UUID")
+    .requiredOption("--to <teamId>", "目标 Team UUID")
+    .action(async (opts) => {
+    const res = await requireClient().reassignAccount(opts.team, opts.account, {
+        toTeamId: opts.to,
+    });
+    console.log(JSON.stringify(res, null, 2));
+});
 // 账号数据多面一次性同步：纯 CLI 编排——把单个 JSON 拆成各写入面，依次调用
 // 已有的 per-resource SDK 方法。服务端无新端点；每个面仍走各自校验过的路由。
 // JSON 中出现的面才写，未出现的跳过；逐面 continue-and-report，任一面失败退出码非零。
@@ -1439,7 +1512,7 @@ drafts
     .description("PATCH /content-drafts/:id")
     .requiredOption("-t, --team <teamId>", "Team UUID")
     .requiredOption("--draft <uuid>", "Draft UUID")
-    .requiredOption("-j, --json <json>", "{ title?, body?, status? }")
+    .requiredOption("-j, --json <json>", "{ title?, body?, status?, expectedVersion? }（传 expectedVersion 开乐观并发,版本不匹配→409)")
     .action(async (opts) => {
     const body = JSON.parse(opts.json);
     const res = await requireClient().updateContentDraft(opts.team, opts.draft, body);
@@ -1950,7 +2023,7 @@ intel
     .command("runs-list")
     .description("GET .../subreddit-intelligence/runs")
     .option("--dimension <d>", "industry|keywords|tier")
-    .option("--status <s>", "queued|running|succeeded|failed")
+    .option("--status <s>", "queued|running|succeeded|partial|failed")
     .option("-n, --limit <n>", "limit", "50")
     .option("--offset <n>", "offset", "0")
     .action(async (opts) => {
@@ -2781,6 +2854,8 @@ reddit
     .option("--repost-type <t>", "repostType filter (none|first|second)")
     .option("--contains-brand-keyword <v>", "true | false | unset")
     .option("--primary-product <text>", "primary product contains match")
+    .option("--sort-by <f>", "score|comments_count|posted_at|view_count|seq_no")
+    .option("--sort-dir <d>", "asc|desc")
     .option("-n, --limit <n>", "limit per page", "100")
     .option("--offset <n>", "pagination offset")
     .option("--all", "fetch all pages (200 per request until total reached)")
@@ -2797,6 +2872,8 @@ reddit
         repostType: opts.repostType,
         containsBrandKeyword: opts.containsBrandKeyword,
         primaryProduct: opts.primaryProduct,
+        sortBy: opts.sortBy,
+        sortDir: opts.sortDir,
         limit: Number(opts.limit),
         offset: opts.offset !== undefined ? Number(opts.offset) : undefined,
         ...redditAggregateQueryParams(opts),