feat(content-studio): 交付即扫即记 + 内容日历只读页（自动化与打磨）

2026-09-13 03:13:41 +08:00 · 2026-07-12 11:51:12 +08:00 · 2026-07-12 11:51:12 +08:00 · ac4b277492
commit ac4b277492
parent f184b94bcd
20 changed files with 647 additions and 120 deletions
--- a/mateclaw-server/src/main/java/vip/mate/content/controller/ContentItemController.java
+++ b/mateclaw-server/src/main/java/vip/mate/content/controller/ContentItemController.java
@ -0,0 +1,46 @@
+package vip.mate.content.controller;
+
+import com.baomidou.mybatisplus.core.metadata.IPage;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import lombok.RequiredArgsConstructor;
+import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RequestParam;
+import org.springframework.web.bind.annotation.RestController;
+import vip.mate.common.result.R;
+import vip.mate.content.model.ContentItemEntity;
+import vip.mate.content.service.ContentItemService;
+
+import java.util.Map;
+
+/**
+ * Read-only content calendar API — lists produced 公众号 / 小红书 pieces and their
+ * lifecycle status, so operators can see what's drafted / packaged / published /
+ * pending. Writes happen through the tools ({@code content_item} + auto-record on
+ * delivery), not here.
+ */
+@Tag(name = "内容日历")
+@RestController
+@RequestMapping("/api/v1/content-items")
+@RequiredArgsConstructor
+public class ContentItemController {
+
+    private final ContentItemService contentItemService;
+
+    @Operation(summary = "内容日历分页列表")
+    @GetMapping
+    public R<IPage<ContentItemEntity>> list(
+            @RequestParam(defaultValue = "1") int page,
+            @RequestParam(defaultValue = "20") int size,
+            @RequestParam(required = false) String platform,
+            @RequestParam(required = false) String status) {
+        return R.ok(contentItemService.page(page, size, platform, status));
+    }
+
+    @Operation(summary = "内容日历状态计数")
+    @GetMapping("/summary")
+    public R<Map<String, Long>> summary() {
+        return R.ok(contentItemService.summary());
+    }
+}
--- a/mateclaw-server/src/main/java/vip/mate/content/service/ContentItemService.java
+++ b/mateclaw-server/src/main/java/vip/mate/content/service/ContentItemService.java
@ -0,0 +1,129 @@
+package vip.mate.content.service;
+
+import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
+import com.baomidou.mybatisplus.core.metadata.IPage;
+import com.baomidou.mybatisplus.extension.plugins.pagination.Page;
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.stereotype.Service;
+import vip.mate.content.model.ContentItemEntity;
+import vip.mate.content.repository.ContentItemMapper;
+
+import java.nio.charset.StandardCharsets;
+import java.security.MessageDigest;
+import java.time.LocalDateTime;
+import java.util.LinkedHashMap;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * Content calendar / dedup ledger service — the single home for content-item
+ * logic, shared by {@code content_item} (the tool), the package tools (which
+ * auto-record on delivery), and the read-only content-calendar API.
+ */
+@Slf4j
+@Service
+@RequiredArgsConstructor
+public class ContentItemService {
+
+    /** Statuses that count as "already covered" for dedup — a discarded draft doesn't. */
+    private static final List<String> COMMITTED_STATUSES = List.of("packaged", "published");
+    /** Ignore same-topic rows created within this window, so a record→check in one
+     *  run doesn't flag itself as a repeat. */
+    private static final long SELF_MATCH_GUARD_MINUTES = 2;
+
+    private final ContentItemMapper contentItemMapper;
+
+    /**
+     * Recent committed items with the same topic fingerprint on this platform,
+     * within {@code days}, excluding just-created rows (self-match guard). Empty
+     * means "not a repeat".
+     */
+    public List<ContentItemEntity> findRecent(String platform, String topic, int days) {
+        LocalDateTime now = LocalDateTime.now();
+        return contentItemMapper.selectList(new LambdaQueryWrapper<ContentItemEntity>()
+                .eq(ContentItemEntity::getPlatform, platform.trim().toLowerCase())
+                .eq(ContentItemEntity::getTopicFingerprint, fingerprint(topic))
+                .in(ContentItemEntity::getStatus, COMMITTED_STATUSES)
+                .ge(ContentItemEntity::getCreateTime, now.minusDays(days))
+                .lt(ContentItemEntity::getCreateTime, now.minusMinutes(SELF_MATCH_GUARD_MINUTES))
+                .orderByDesc(ContentItemEntity::getCreateTime));
+    }
+
+    /** Record a produced piece; returns the new item id. */
+    public Long record(Long workspaceId, String platform, String topic, String title,
+                       String status, String previewUrl, String externalRef) {
+        ContentItemEntity e = new ContentItemEntity();
+        e.setWorkspaceId(workspaceId);
+        e.setPlatform(platform.trim().toLowerCase());
+        e.setTopic(topic != null ? topic.trim() : null);
+        e.setTopicFingerprint(fingerprint(topic != null ? topic : title));
+        e.setTitle(title != null ? title.trim() : null);
+        e.setStatus(status == null || status.isBlank() ? "packaged" : status.trim().toLowerCase());
+        e.setPreviewUrl(previewUrl);
+        e.setExternalRef(externalRef);
+        contentItemMapper.insert(e);
+        log.info("[ContentItem] recorded id={} ws={} platform={} status={} title='{}'",
+                e.getId(), workspaceId, e.getPlatform(), e.getStatus(), title);
+        return e.getId();
+    }
+
+    /** Flip an item to published. Returns false if the id is unknown. */
+    public boolean markPublished(Long id, String externalRef) {
+        ContentItemEntity e = contentItemMapper.selectById(id);
+        if (e == null) {
+            return false;
+        }
+        e.setStatus("published");
+        e.setPublishTime(LocalDateTime.now());
+        if (externalRef != null && !externalRef.isBlank()) {
+            e.setExternalRef(externalRef);
+        }
+        contentItemMapper.updateById(e);
+        log.info("[ContentItem] item {} marked published (ref={})", id, externalRef);
+        return true;
+    }
+
+    /** Paged content-calendar listing, newest first, optionally filtered by platform / status. */
+    public IPage<ContentItemEntity> page(int page, int size, String platform, String status) {
+        LambdaQueryWrapper<ContentItemEntity> w = new LambdaQueryWrapper<>();
+        if (platform != null && !platform.isBlank()) {
+            w.eq(ContentItemEntity::getPlatform, platform.trim().toLowerCase());
+        }
+        if (status != null && !status.isBlank()) {
+            w.eq(ContentItemEntity::getStatus, status.trim().toLowerCase());
+        }
+        w.orderByDesc(ContentItemEntity::getCreateTime);
+        int p = Math.max(1, page);
+        int s = Math.min(Math.max(1, size), 100);
+        return contentItemMapper.selectPage(new Page<>(p, s), w);
+    }
+
+    /** Counts by status (draft/packaged/published/failed) plus total, for the summary cards. */
+    public Map<String, Long> summary() {
+        Map<String, Long> m = new LinkedHashMap<>();
+        for (String s : List.of("draft", "packaged", "published", "failed")) {
+            m.put(s, contentItemMapper.selectCount(
+                    new LambdaQueryWrapper<ContentItemEntity>().eq(ContentItemEntity::getStatus, s)));
+        }
+        m.put("total", contentItemMapper.selectCount(null));
+        return m;
+    }
+
+    /** Stable 32-hex fingerprint of the normalized topic (lowercased, alnum/CJK only). */
+    public static String fingerprint(String topic) {
+        String normalized = topic == null ? "" : topic.toLowerCase()
+                .replaceAll("[\\s\\p{Punct}\\u3000-\\u303F\\uFF00-\\uFFEF]+", "");
+        try {
+            byte[] hash = MessageDigest.getInstance("SHA-256")
+                    .digest(normalized.getBytes(StandardCharsets.UTF_8));
+            StringBuilder hex = new StringBuilder();
+            for (int i = 0; i < 16; i++) {
+                hex.append(String.format("%02x", hash[i]));
+            }
+            return hex.toString();
+        } catch (Exception e) {
+            return Integer.toHexString(normalized.hashCode());
+        }
+    }
+}
--- a/mateclaw-server/src/main/java/vip/mate/tool/builtin/ComplianceScanTool.java
+++ b/mateclaw-server/src/main/java/vip/mate/tool/builtin/ComplianceScanTool.java
@ -5,6 +5,8 @@ import org.springframework.ai.tool.annotation.Tool;
 import org.springframework.ai.tool.annotation.ToolParam;
 import org.springframework.stereotype.Component;

+import java.util.Arrays;
+
 /**
 * Built-in tool: server-side compliance scan for 公众号 / 小红书 copy. Deterministic
 * backstop for the model's skill-side self-check — catches 广告法 极限词, WeChat 诱导
@ -25,8 +27,12 @@ public class ComplianceScanTool {
        """)
    public String compliance_scan(
            @ToolParam(description = "Text to scan (title + body)")
-            String text) {
-        ComplianceScanner.Result result = ComplianceScanner.scan(text);
+            String text,
+            @ToolParam(description = "Extra banned words to enforce, comma-separated (e.g. recalled banned_words)", required = false)
+            String extraBannedWords) {
+        ComplianceScanner.Result result = (extraBannedWords == null || extraBannedWords.isBlank())
+                ? ComplianceScanner.scan(text)
+                : ComplianceScanner.scan(text, Arrays.asList(extraBannedWords.split(",")));
        log.info("[ComplianceScan] hits={}, highRisk={}", result.hits().size(), result.hasHighRisk());
        return ComplianceScanner.report(result);
    }
--- a/mateclaw-server/src/main/java/vip/mate/tool/builtin/ComplianceScanner.java
+++ b/mateclaw-server/src/main/java/vip/mate/tool/builtin/ComplianceScanner.java
@ -1,6 +1,7 @@
 package vip.mate.tool.builtin;

 import java.util.ArrayList;
+import java.util.Collection;
 import java.util.LinkedHashMap;
 import java.util.List;
 import java.util.Map;
@ -83,6 +84,34 @@ final class ComplianceScanner {
        return new Result(hits);
    }

+    /**
+     * Scan with additional user-supplied banned words merged in as a
+     * (non-high-risk) {@code 自定义禁用词} category — e.g. the user's
+     * {@code banned_words} memory or brand-forbidden terms.
+     */
+    static Result scan(String text, Collection<String> extraTerms) {
+        Result base = scan(text);
+        if (extraTerms == null || extraTerms.isEmpty() || text == null || text.isBlank()) {
+            return base;
+        }
+        List<String> hitTerms = new ArrayList<>();
+        for (String t : extraTerms) {
+            if (t == null) {
+                continue;
+            }
+            String term = t.trim();
+            if (!term.isEmpty() && text.contains(term) && !hitTerms.contains(term)) {
+                hitTerms.add(term);
+            }
+        }
+        if (hitTerms.isEmpty()) {
+            return base;
+        }
+        List<CategoryHit> all = new ArrayList<>(base.hits());
+        all.add(new CategoryHit("自定义禁用词", hitTerms, false));
+        return new Result(all);
+    }
+
    /** Render a scan result as a short Chinese report. */
    static String report(Result result) {
        if (result.clean()) {
--- a/mateclaw-server/src/main/java/vip/mate/tool/builtin/ContentItemTool.java
+++ b/mateclaw-server/src/main/java/vip/mate/tool/builtin/ContentItemTool.java
@ -1,31 +1,30 @@
 package vip.mate.tool.builtin;

-import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
+import org.springframework.ai.chat.model.ToolContext;
 import org.springframework.ai.tool.annotation.Tool;
 import org.springframework.ai.tool.annotation.ToolParam;
+import org.springframework.lang.Nullable;
 import org.springframework.stereotype.Component;
+import vip.mate.agent.context.ChatOrigin;
 import vip.mate.content.model.ContentItemEntity;
-import vip.mate.content.repository.ContentItemMapper;
+import vip.mate.content.service.ContentItemService;

-import java.nio.charset.StandardCharsets;
-import java.security.MessageDigest;
-import java.time.LocalDateTime;
 import java.util.List;

 /**
 * Built-in tool: the content calendar / dedup ledger for the content studio.
- * Lets the daily scheduler avoid repeating a topic, records produced pieces, and
- * marks them published so publishing is idempotent and auditable.
+ * Thin wrapper over {@link ContentItemService}; the package tools also call the
+ * service to auto-record on delivery.
 *
 * <p>Actions:
 * <ul>
 *   <li>{@code check_recent} — has this topic been produced on this platform in
 *       the last N days? Call BEFORE committing to a topic.</li>
- *   <li>{@code record} — log a produced piece (draft/packaged) with its title and
- *       preview link.</li>
- *   <li>{@code mark_published} — flip an item to published with its platform ref.</li>
+ *   <li>{@code record} — log a produced piece (usually done automatically by the
+ *       package tools; available for manual use).</li>
+ *   <li>{@code mark_published} — flip an item to published.</li>
 * </ul>
 */
@Slf4j
@ -35,7 +34,7 @@ public class ContentItemTool {

    private static final int DEFAULT_RECENT_DAYS = 14;

-    private final ContentItemMapper contentItemMapper;
+    private final ContentItemService contentItemService;

    @Tool(name = "content_item", description = """
        Content calendar / dedup ledger for 公众号 & 小红书 pieces.
@ -43,13 +42,12 @@ public class ContentItemTool {
        Actions:
          - check_recent: has `topic` already been produced for `platform`
            (gzh|xhs) within the last `days` (default 14)? Call this BEFORE picking
-            a topic in a scheduled run, to avoid repeats. Returns whether it's a
-            repeat plus recent titles.
-          - record: log a produced piece — platform, topic, title, status
-            (draft|packaged|published, default packaged), optional previewUrl /
-            externalRef. Returns the item id.
-          - mark_published: set item `id` to published with optional externalRef
-            (draft media_id / publish id).
+            a topic in a scheduled run, to avoid repeats. Only counts committed
+            pieces (packaged/published). Returns whether it's a repeat plus titles.
+          - record: log a produced piece (platform, topic, title, status). NOTE:
+            gzh_package / xhs_package already auto-record on delivery, so you rarely
+            need this by hand.
+          - mark_published: set item `id` to published with optional externalRef.
        """)
    public String content_item(
            @ToolParam(description = "Action: check_recent | record | mark_published")
@ -69,12 +67,13 @@ public class ContentItemTool {
            @ToolParam(description = "Recency window in days for check_recent (default 14)", required = false)
            Integer days,
            @ToolParam(description = "Item id (mark_published)", required = false)
-            Long id) {
+            Long id,
+            @Nullable ToolContext ctx) {

        String act = action == null ? "" : action.trim().toLowerCase();
        return switch (act) {
            case "check_recent" -> checkRecent(platform, topic, days);
-            case "record" -> record(platform, topic, title, status, previewUrl, externalRef);
+            case "record" -> record(platform, topic, title, status, previewUrl, externalRef, ctx);
            case "mark_published" -> markPublished(id, externalRef);
            default -> "Error: unknown action '" + act + "'. Use check_recent | record | mark_published.";
        };
@ -85,14 +84,7 @@ public class ContentItemTool {
            return "Error: platform and topic are required for check_recent.";
        }
        int window = (days == null || days <= 0) ? DEFAULT_RECENT_DAYS : days;
-        LocalDateTime cutoff = LocalDateTime.now().minusDays(window);
-        String fp = fingerprint(topic);
-        List<ContentItemEntity> recent = contentItemMapper.selectList(
-                new LambdaQueryWrapper<ContentItemEntity>()
-                        .eq(ContentItemEntity::getPlatform, platform.trim().toLowerCase())
-                        .eq(ContentItemEntity::getTopicFingerprint, fp)
-                        .ge(ContentItemEntity::getCreateTime, cutoff)
-                        .orderByDesc(ContentItemEntity::getCreateTime));
+        List<ContentItemEntity> recent = contentItemService.findRecent(platform, topic, window);
        if (recent.isEmpty()) {
            return "✅ 未重复：最近 " + window + " 天没有在 " + platform + " 做过「" + topic + "」，可以继续。";
        }
@ -109,57 +101,27 @@ public class ContentItemTool {
    }

    private String record(String platform, String topic, String title, String status,
-                          String previewUrl, String externalRef) {
+                          String previewUrl, String externalRef, @Nullable ToolContext ctx) {
        if (isBlank(platform) || isBlank(topic)) {
            return "Error: platform and topic are required for record.";
        }
-        ContentItemEntity e = new ContentItemEntity();
-        e.setPlatform(platform.trim().toLowerCase());
-        e.setTopic(topic.trim());
-        e.setTopicFingerprint(fingerprint(topic));
-        e.setTitle(title != null ? title.trim() : null);
-        e.setStatus(isBlank(status) ? "packaged" : status.trim().toLowerCase());
-        e.setPreviewUrl(previewUrl);
-        e.setExternalRef(externalRef);
-        contentItemMapper.insert(e);
-        log.info("[ContentItem] recorded id={} platform={} status={} title='{}'",
-                e.getId(), e.getPlatform(), e.getStatus(), title);
-        return "✅ 已记入内容日历。item id: " + e.getId() + "（status=" + e.getStatus() + "）";
+        Long id = contentItemService.record(workspaceFromContext(ctx), platform, topic, title,
+                status, previewUrl, externalRef);
+        return "✅ 已记入内容日历。item id: " + id + "（status=" + (isBlank(status) ? "packaged" : status) + "）";
    }

    private String markPublished(Long id, String externalRef) {
        if (id == null) {
            return "Error: id is required for mark_published.";
        }
-        ContentItemEntity e = contentItemMapper.selectById(id);
-        if (e == null) {
-            return "Error: content item " + id + " not found.";
-        }
-        e.setStatus("published");
-        e.setPublishTime(LocalDateTime.now());
-        if (!isBlank(externalRef)) {
-            e.setExternalRef(externalRef);
-        }
-        contentItemMapper.updateById(e);
-        log.info("[ContentItem] item {} marked published (ref={})", id, externalRef);
-        return "✅ 已标记为已发布。item id: " + id;
+        return contentItemService.markPublished(id, externalRef)
+                ? "✅ 已标记为已发布。item id: " + id
+                : "Error: content item " + id + " not found.";
    }

-    /** Stable 32-hex fingerprint of the normalized topic (lowercased, alnum/CJK only). */
-    static String fingerprint(String topic) {
-        String normalized = topic == null ? "" : topic.toLowerCase()
-                .replaceAll("[\\s\\p{Punct}\\u3000-\\u303F\\uFF00-\\uFFEF]+", "");
-        try {
-            byte[] hash = MessageDigest.getInstance("SHA-256")
-                    .digest(normalized.getBytes(StandardCharsets.UTF_8));
-            StringBuilder hex = new StringBuilder();
-            for (int i = 0; i < 16; i++) {
-                hex.append(String.format("%02x", hash[i]));
-            }
-            return hex.toString();
-        } catch (Exception e) {
-            return Integer.toHexString(normalized.hashCode());
-        }
+    private static Long workspaceFromContext(@Nullable ToolContext ctx) {
+        ChatOrigin origin = ChatOrigin.from(ctx);
+        return origin != null && origin.workspaceId() != null ? origin.workspaceId() : 1L;
    }

    private static boolean isBlank(String s) {
--- a/mateclaw-server/src/main/java/vip/mate/tool/builtin/GzhPackageTool.java
+++ b/mateclaw-server/src/main/java/vip/mate/tool/builtin/GzhPackageTool.java
@ -15,6 +15,8 @@ import org.springframework.ai.tool.annotation.Tool;
 import org.springframework.ai.tool.annotation.ToolParam;
 import org.springframework.lang.Nullable;
 import org.springframework.stereotype.Component;
+import vip.mate.agent.context.ChatOrigin;
+import vip.mate.content.service.ContentItemService;
 import vip.mate.tool.browser.UrlSafetyChecker;
 import vip.mate.tool.document.GeneratedFileCache;

@ -69,6 +71,7 @@ public class GzhPackageTool {
    private static final String CODE_BG = "#f6f8fa";

    private final GeneratedFileCache cache;
+    private final ContentItemService contentItemService;

    @Tool(name = "gzh_package", description = """
        Package a finished WeChat Official Account (公众号) article and return an
@ -96,6 +99,8 @@ public class GzhPackageTool {
            String coverImageUrl,
            @ToolParam(description = "Author / source name", required = false)
            String author,
+            @ToolParam(description = "Selected topic (for the content ledger; falls back to title)", required = false)
+            String topic,
            @Nullable ToolContext ctx) {

        if (title == null || title.isBlank()) {
@ -181,9 +186,25 @@ public class GzhPackageTool {
            out.append("📦 素材下载（article.html + article.md + 封面，").append(coverNote).append("）：")
               .append(zipUrl).append('\n');
        }
+        // Auto compliance scan on delivery — never relies on the model calling it.
+        ComplianceScanner.Result scan = ComplianceScanner.scan(title + "\n" + markdown);
+        if (!scan.clean()) {
+            out.append('\n').append(ComplianceScanner.report(scan)).append('\n');
+        }
        out.append("\n可将下面的内联样式 HTML 直接粘贴进公众号编辑器（如需直接进草稿箱，用 gzh_publish）：\n");
        out.append("```html\n").append(container).append("\n```");
-        log.info("[GzhPackage] packaged '{}' ({} md chars, coverResolved={})", title, markdown.length(), cover != null);
+
+        // Auto-record into the content ledger — the calendar is always populated.
+        try {
+            Long itemId = contentItemService.record(workspaceFromContext(ctx), "gzh",
+                    topic != null && !topic.isBlank() ? topic : title.trim(),
+                    title.trim(), "packaged", previewUrl, null);
+            out.append("\n🗓️ 已记入内容日历（item id: ").append(itemId).append("）。");
+        } catch (Exception e) {
+            log.warn("[GzhPackage] auto-record failed: {}", e.getMessage());
+        }
+        log.info("[GzhPackage] packaged '{}' ({} md chars, coverResolved={}, complianceHits={})",
+                title, markdown.length(), cover != null, scan.hits().size());
        return out.toString();
    }

@ -332,6 +353,11 @@ public class GzhPackageTool {
        }
    }

+    private static Long workspaceFromContext(@Nullable ToolContext ctx) {
+        ChatOrigin origin = ChatOrigin.from(ctx);
+        return origin != null && origin.workspaceId() != null ? origin.workspaceId() : 1L;
+    }
+
    private static boolean isImage(GeneratedFileCache.Entry e) {
        return e.mimeType() != null && e.mimeType().startsWith("image/");
    }
--- a/mateclaw-server/src/main/java/vip/mate/tool/builtin/XhsPackageTool.java
+++ b/mateclaw-server/src/main/java/vip/mate/tool/builtin/XhsPackageTool.java
@ -8,6 +8,8 @@ import org.springframework.ai.tool.annotation.Tool;
 import org.springframework.ai.tool.annotation.ToolParam;
 import org.springframework.lang.Nullable;
 import org.springframework.stereotype.Component;
+import vip.mate.agent.context.ChatOrigin;
+import vip.mate.content.service.ContentItemService;
 import vip.mate.tool.browser.UrlSafetyChecker;
 import vip.mate.tool.document.GeneratedFileCache;
 import vip.mate.tool.guard.WorkspacePathGuard;
@ -66,6 +68,7 @@ public class XhsPackageTool {
    private static final String PAGE_BG = "#f4f4f4";

    private final GeneratedFileCache cache;
+    private final ContentItemService contentItemService;

    @Tool(name = "xhs_package", description = """
        Package a Xiaohongshu (小红书) note into an IMAGE-FIRST online preview plus a
@ -94,6 +97,8 @@ public class XhsPackageTool {
            String tags,
            @ToolParam(description = "Comma-separated image references in display order (first = cover); >=3 required")
            String images,
+            @ToolParam(description = "Selected topic (for the content ledger; falls back to title)", required = false)
+            String topic,
            @Nullable ToolContext ctx) {

        if (title == null || title.isBlank()) {
@ -156,8 +161,23 @@ public class XhsPackageTool {
        if (!skipped.isEmpty()) {
            out.append("⚠️ 未打包：").append(String.join("；", skipped)).append('\n');
        }
+        // Auto compliance scan on delivery — never relies on the model calling it.
+        ComplianceScanner.Result scan = ComplianceScanner.scan(title + "\n" + (body == null ? "" : body));
+        if (!scan.clean()) {
+            out.append(ComplianceScanner.report(scan)).append('\n');
+        }
+        // Auto-record into the content ledger.
+        try {
+            Long itemId = contentItemService.record(workspaceFromContext(ctx), "xhs",
+                    topic != null && !topic.isBlank() ? topic : title.trim(),
+                    title.trim(), "packaged", previewUrl, null);
+            out.append("🗓️ 已记入内容日历（item id: ").append(itemId).append("）。\n");
+        } catch (Exception e) {
+            log.warn("[XhsPackage] auto-record failed: {}", e.getMessage());
+        }
        out.append('\n').append(guideText());
-        log.info("[XhsPackage] packaged '{}' ({} images, {} skipped)", title, imgs.size(), skipped.size());
+        log.info("[XhsPackage] packaged '{}' ({} images, {} skipped, complianceHits={})",
+                title, imgs.size(), skipped.size(), scan.hits().size());
        return out.toString();
    }

@ -309,6 +329,11 @@ public class XhsPackageTool {
            5. 核对无违禁词后自行发布。""".formatted(CREATOR_URL);
    }

+    private static Long workspaceFromContext(@Nullable ToolContext ctx) {
+        ChatOrigin origin = ChatOrigin.from(ctx);
+        return origin != null && origin.workspaceId() != null ? origin.workspaceId() : 1L;
+    }
+
    /** Last path segment of a reference, minus any {@code ?query} / {@code #fragment}. */
    private static String lastSegment(String ref) {
        String s = ref.split("[?#]")[0];
--- a/mateclaw-server/src/main/resources/skills/gzh_article/SKILL.md
+++ b/mateclaw-server/src/main/resources/skills/gzh_article/SKILL.md
@ -1,7 +1,7 @@
 ---
 name: gzh_article
 description: '公众号图文创作 / 推文 / 官方号文章 (official account article) — 端到端：选题→搜集→成文→配图→去AI化→公众号内联样式排版→交付/草稿箱。honors user persona & style memory.'
-version: 1.2.0
+version: 1.3.0
 tags:
 - 公众号
 - 图文
@ -102,13 +102,13 @@ gzh_package(title="<标题>", markdown="<正文 Markdown，含小标题/列表/

 `references/` 里的 `gzh_layout*.html` 是**服务端配色/排版的参考**（`gzh_package` 已内置同款风格）；只有当用户明确要**高度定制的特殊版式**、且篇幅不大时，才手写内联 HTML 并用 `render_html_image(html=...)` 出预览图，注意控制体量避免截断。

-## 定时 / 批量场景：内容日历 + 合规硬闸
+## 定时 / 批量场景：内容日历 + 合规

-长期投产（尤其每日定时）时，多接两步，避免重复选题和违规发布：
+长期投产（尤其每日定时）时：

- **选题前查重**：`content_item(action="check_recent", platform="gzh", topic="<选题>")`。命中"疑似重复"就换角度或另选。
- **发布前硬扫**：把标题+正文交给 `compliance_scan` 服务端扫一遍（极限词/诱导/承诺/功效）。命中高危词先替换——`gzh_publish` 进草稿箱时也会对高危词硬拦截。
- **产出落台账**：`gzh_package` / `gzh_publish` 完成后 `content_item(action="record", platform="gzh", title, previewUrl, status="packaged"|"draft")`；真正发表后 `content_item(action="mark_published", id, externalRef)`。让每日不重题、发布可追溯。
+- **选题前查重（务必在成文前）**：`content_item(action="check_recent", platform="gzh", topic="<选题>")`。命中"疑似重复"就换角度或另选。只计已打包/已发布，不会把你这一轮刚产出的算进去。
+- **交付即扫即记（自动，无需手动）**：`gzh_package` 交付时会**自动**跑合规扫描（极限词/诱导/承诺/功效，报告并入返回）并**自动记入内容日历**。你只需在 `gzh_package` 里传上 `topic="<选题>"`（与 `title` 区分，用于台账去重指纹）。个人 / 品牌禁用词可另调 `compliance_scan(text, extraBannedWords="<recall 到的 banned_words，逗号分隔>")`。
+- **发布**：`gzh_publish` 进草稿箱时对高危词（极限/诱导/承诺）硬拦截；真正发表后 `content_item(action="mark_published", id, externalRef)`。

 ## 保存自定义模板 / 对话升级技能

--- a/mateclaw-server/src/main/resources/skills/xhs_note/SKILL.md
+++ b/mateclaw-server/src/main/resources/skills/xhs_note/SKILL.md
@ -1,7 +1,7 @@
 ---
 name: xhs_note
 description: '小红书图文创作 / 笔记 / 种草文案 (xiaohongshu / red note) — 端到端：成文→配图(≥3 张竖版)→去AI化→在线预览打包交付。以图为主、文字辅助：标题四件套 + 碎句正文 + 话题标签，配 3:4 竖版卡片，最少 3 张图。honors user persona & style memory.'
-version: 1.2.0
+version: 1.3.0
 tags:
 - 小红书
 - 图文
@ -118,11 +118,11 @@ xhs_package(title="<标题>", body="<正文，含 emoji 与换行>",

 ## 定时 / 批量场景：内容日历 + 合规

-长期投产（每日定时）时多接两步：
+长期投产（每日定时）时：

- **选题前查重**：`content_item(action="check_recent", platform="xhs", topic="<选题>")`，命中就换角度。
- **打包前硬扫**：把标题+正文交给 `compliance_scan`（极限词/诱导/承诺/功效），命中先替换。
- **产出落台账**：`xhs_package` 完成后 `content_item(action="record", platform="xhs", title, previewUrl, status="packaged")`；用户手动上传发布后 `content_item(action="mark_published", id)`。这样能知道哪些已发、哪些还在待办。
+- **选题前查重（成文前）**：`content_item(action="check_recent", platform="xhs", topic="<选题>")`，命中就换角度。只计已打包/已发布。
+- **交付即扫即记（自动）**：`xhs_package` 交付时会**自动**跑合规扫描并**自动记入内容日历**——你只需在 `xhs_package` 里传上 `topic="<选题>"`（与 `title` 区分）。个人 / 品牌禁用词可另调 `compliance_scan(text, extraBannedWords="<recall 到的 banned_words>")`。
+- **上传后**：用户手动发布后 `content_item(action="mark_published", id)`，就能知道哪些已发、哪些还在待办。

 ## 保存自定义卡片模板 / 对话升级技能

--- a/mateclaw-server/src/test/java/vip/mate/tool/builtin/ComplianceScannerTest.java
+++ b/mateclaw-server/src/test/java/vip/mate/tool/builtin/ComplianceScannerTest.java
@ -3,6 +3,8 @@ package vip.mate.tool.builtin;
 import org.junit.jupiter.api.DisplayName;
 import org.junit.jupiter.api.Test;

+import java.util.List;
+
 import static org.junit.jupiter.api.Assertions.*;

 /**
@ -43,6 +45,19 @@ class ComplianceScannerTest {
        assertFalse(r.hasHighRisk(), "医疗功效 is a warning, not a hard block");
    }

+    @Test
+    @DisplayName("extra banned words merge in as a non-high-risk 自定义禁用词 category")
+    void extraBannedWords() {
+        ComplianceScanner.Result r = ComplianceScanner.scan(
+                "这段文字提到了竞品X和内部代号Y", List.of("竞品X", "内部代号Y", "没出现的词"));
+        assertFalse(r.clean());
+        assertFalse(r.hasHighRisk(), "custom banned words are a warning, not a hard block");
+        String rep = ComplianceScanner.report(r);
+        assertTrue(rep.contains("自定义禁用词"));
+        assertTrue(rep.contains("竞品X") && rep.contains("内部代号Y"));
+        assertFalse(rep.contains("没出现的词"), "only actually-present terms are reported");
+    }
+
    @Test
    @DisplayName("clean copy scans clean")
    void cleanCopy() {
--- a/mateclaw-server/src/test/java/vip/mate/tool/builtin/ContentItemToolTest.java
+++ b/mateclaw-server/src/test/java/vip/mate/tool/builtin/ContentItemToolTest.java
@ -1,50 +1,48 @@
 package vip.mate.tool.builtin;

-import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.DisplayName;
 import org.junit.jupiter.api.Test;
 import vip.mate.content.model.ContentItemEntity;
-import vip.mate.content.repository.ContentItemMapper;
+import vip.mate.content.service.ContentItemService;

 import java.time.LocalDateTime;
 import java.util.List;

 import static org.junit.jupiter.api.Assertions.*;
-import static org.mockito.ArgumentMatchers.any;
+import static org.mockito.ArgumentMatchers.*;
 import static org.mockito.Mockito.*;

 /**
- * Pin {@link ContentItemTool}: the topic fingerprint is stable across cosmetic
- * differences (so repeats are caught), and the check_recent / record /
- * mark_published actions behave.
+ * Pin {@link ContentItemTool}: the fingerprint is stable across cosmetic
+ * differences, and check_recent / record / mark_published delegate correctly.
 */
 class ContentItemToolTest {

-    private ContentItemMapper mapper;
+    private ContentItemService service;
    private ContentItemTool tool;

    @BeforeEach
    void setUp() {
-        mapper = mock(ContentItemMapper.class);
-        tool = new ContentItemTool(mapper);
+        service = mock(ContentItemService.class);
+        tool = new ContentItemTool(service);
    }

    @Test
    @DisplayName("fingerprint ignores case / whitespace / punctuation but distinguishes real topics")
    void fingerprintStable() {
-        String a = ContentItemTool.fingerprint("周末咖啡探店");
-        String b = ContentItemTool.fingerprint("  周末 咖啡，探店！ ");
+        String a = ContentItemService.fingerprint("周末咖啡探店");
+        String b = ContentItemService.fingerprint("  周末 咖啡，探店！ ");
        assertEquals(a, b, "cosmetic differences must collapse to the same fingerprint");
-        assertNotEquals(a, ContentItemTool.fingerprint("露营装备清单"), "different topics differ");
+        assertNotEquals(a, ContentItemService.fingerprint("露营装备清单"), "different topics differ");
    }

    @Test
    @DisplayName("check_recent: empty history → not a repeat")
    void checkRecentEmpty() {
-        when(mapper.selectList(any())).thenReturn(List.of());
+        when(service.findRecent(eq("gzh"), eq("周末咖啡探店"), anyInt())).thenReturn(List.of());
        String out = tool.content_item("check_recent", "gzh", "周末咖啡探店",
-                null, null, null, null, 14, null);
+                null, null, null, null, 14, null, null);
        assertTrue(out.contains("未重复"), out);
    }

@ -55,42 +53,42 @@ class ContentItemToolTest {
        prior.setTitle("上周那篇咖啡探店");
        prior.setStatus("published");
        prior.setCreateTime(LocalDateTime.now().minusDays(3));
-        when(mapper.selectList(any())).thenReturn(List.of(prior));
+        when(service.findRecent(eq("gzh"), eq("周末咖啡探店"), anyInt())).thenReturn(List.of(prior));

        String out = tool.content_item("check_recent", "gzh", "周末咖啡探店",
-                null, null, null, null, 14, null);
+                null, null, null, null, 14, null, null);
        assertTrue(out.contains("疑似重复"), out);
        assertTrue(out.contains("上周那篇咖啡探店"), "should show the prior title");
    }

    @Test
-    @DisplayName("record: inserts a row and reports the item")
-    void recordInserts() {
+    @DisplayName("record: delegates to the service and reports the item id")
+    void recordDelegates() {
+        when(service.record(any(), eq("xhs"), eq("露营装备清单"), any(), any(), any(), any()))
+                .thenReturn(999L);
        String out = tool.content_item("record", "xhs", "露营装备清单",
-                "新手露营必带的8样东西", "packaged", "http://x/preview", null, null, null);
-        verify(mapper, times(1)).insert(any(ContentItemEntity.class));
-        assertTrue(out.contains("已记入内容日历"), out);
+                "新手露营必带的8样东西", "packaged", "http://x/preview", null, null, null, null);
+        verify(service, times(1)).record(any(), eq("xhs"), eq("露营装备清单"),
+                eq("新手露营必带的8样东西"), eq("packaged"), eq("http://x/preview"), isNull());
+        assertTrue(out.contains("999"), out);
    }

    @Test
-    @DisplayName("mark_published: flips status and stamps publish time")
+    @DisplayName("mark_published: reports success / not-found from the service")
    void markPublished() {
-        ContentItemEntity e = new ContentItemEntity();
-        e.setStatus("packaged");
-        when(mapper.selectById(123L)).thenReturn(e);
+        when(service.markPublished(123L, "media_abc")).thenReturn(true);
+        assertTrue(tool.content_item("mark_published", null, null, null, null,
+                null, "media_abc", null, 123L, null).contains("已标记为已发布"));

-        String out = tool.content_item("mark_published", null, null, null, null,
-                null, "media_abc", null, 123L);
-        assertTrue(out.contains("已标记为已发布"), out);
-        assertEquals("published", e.getStatus());
-        assertNotNull(e.getPublishTime());
-        verify(mapper).updateById(e);
+        when(service.markPublished(404L, null)).thenReturn(false);
+        assertTrue(tool.content_item("mark_published", null, null, null, null,
+                null, null, null, 404L, null).startsWith("Error:"));
    }

    @Test
    @DisplayName("unknown action is rejected")
    void unknownAction() {
-        assertTrue(tool.content_item("frobnicate", null, null, null, null, null, null, null, null)
+        assertTrue(tool.content_item("frobnicate", null, null, null, null, null, null, null, null, null)
                .startsWith("Error:"));
    }
 }
--- a/mateclaw-server/src/test/java/vip/mate/tool/builtin/GzhPackageCoverHealingTest.java
+++ b/mateclaw-server/src/test/java/vip/mate/tool/builtin/GzhPackageCoverHealingTest.java
@ -4,11 +4,14 @@ import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.DisplayName;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.io.TempDir;
+import vip.mate.content.service.ContentItemService;
 import vip.mate.tool.document.GeneratedFileCache;

 import java.nio.file.Path;

 import static org.junit.jupiter.api.Assertions.*;
+import static org.mockito.ArgumentMatchers.*;
+import static org.mockito.Mockito.*;

 /**
 * Reproduce and pin the fix for the broken-cover bug: when the model references
@ -22,12 +25,14 @@ import static org.junit.jupiter.api.Assertions.*;
 class GzhPackageCoverHealingTest {

    private GeneratedFileCache cache;
+    private ContentItemService contentSvc;
    private GzhPackageTool tool;

    @BeforeEach
    void setUp(@TempDir Path tempDir) {
        cache = new GeneratedFileCache(tempDir);
-        tool = new GzhPackageTool(cache);
+        contentSvc = mock(ContentItemService.class);
+        tool = new GzhPackageTool(cache, contentSvc);
    }

    private static final String BODY = "## 小节一\n\n正文一段。\n\n## 小节二\n\n又一段。";
@ -42,6 +47,7 @@ class GzhPackageCoverHealingTest {
                BODY,
                "/api/v1/files/generated/cover_cat_7pits.png",
                "内容工作室",
+                null,
                null);

        assertTrue(out.contains("/api/v1/files/generated/" + id),
@ -60,6 +66,7 @@ class GzhPackageCoverHealingTest {
                BODY,
                "/api/v1/files/generated/deadbeef-0000-0000-0000-000000000000",
                "内容工作室",
+                null,
                null);

        assertTrue(out.contains("⚠️"), "an unresolved cover must be flagged; got:\n" + out);
@ -77,6 +84,7 @@ class GzhPackageCoverHealingTest {
                BODY,
                "/api/v1/files/generated/" + id,
                "内容工作室",
+                null,
                null);

        assertTrue(out.contains("/api/v1/files/generated/" + id), "the cover id must be embedded");
@ -84,6 +92,27 @@ class GzhPackageCoverHealingTest {
        assertFalse(out.contains("⚠️"), "a resolved cover must not warn");
    }

+    @Test
+    @DisplayName("delivery auto-scans compliance and auto-records to the ledger")
+    void autoScanAndRecordOnDelivery() {
+        String id = cache.put("PNGBYTES".getBytes(), "cover.png", "image/png");
+        String out = tool.gzh_package(
+                "夏日凉拌菜",
+                "## 小节\n本店全国第一、集赞20个送好礼。",  // 极限词 + 诱导 → 高危命中
+                "/api/v1/files/generated/" + id,
+                "内容工作室",
+                "夏天凉拌菜",  // topic distinct from title
+                null);
+
+        // Compliance report is baked into the delivery output (not left to the model).
+        assertTrue(out.contains("合规扫描命中"), "auto-scan report must be in the output:\n" + out);
+        assertTrue(out.contains("广告法极限词") || out.contains("微信诱导"), "should flag the violation");
+        // Auto-record uses the topic (not the title) for the ledger fingerprint.
+        assertTrue(out.contains("已记入内容日历"), "auto-record note must appear");
+        verify(contentSvc, times(1)).record(any(), eq("gzh"), eq("夏天凉拌菜"),
+                eq("夏日凉拌菜"), eq("packaged"), any(), isNull());
+    }
+
    @Test
    @DisplayName("a non-image generated file referenced as cover → placeholder, not the non-image")
    void nonImageReferenceFallsBackToPlaceholder() {
@ -93,6 +122,7 @@ class GzhPackageCoverHealingTest {
                BODY,
                "/api/v1/files/generated/" + id,
                "内容工作室",
+                null,
                null);

        assertTrue(out.contains("⚠️"), "a non-image cover must be flagged");
--- a/mateclaw-server/src/test/java/vip/mate/tool/builtin/XhsPackageTest.java
+++ b/mateclaw-server/src/test/java/vip/mate/tool/builtin/XhsPackageTest.java
@ -4,8 +4,11 @@ import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.DisplayName;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.io.TempDir;
+import vip.mate.content.service.ContentItemService;
 import vip.mate.tool.document.GeneratedFileCache;

+import static org.mockito.Mockito.mock;
+
 import java.nio.charset.StandardCharsets;
 import java.nio.file.Path;
 import java.util.regex.Matcher;
@ -26,7 +29,7 @@ class XhsPackageTest {
    @BeforeEach
    void setUp(@TempDir Path tempDir) {
        cache = new GeneratedFileCache(tempDir);
-        tool = new XhsPackageTool(cache);
+        tool = new XhsPackageTool(cache, mock(ContentItemService.class));
    }

    private String putImg(String name) {
@ -45,7 +48,7 @@ class XhsPackageTest {
    @DisplayName("fewer than 3 images → refused, no preview minted")
    void refusesUnderThreeImages() {
        String imgs = putImg("cover.png") + "," + putImg("c1.png");
-        String out = tool.xhs_package("夏日穿搭", "正文", "穿搭,夏天", imgs, null);
+        String out = tool.xhs_package("夏日穿搭", "正文", "穿搭,夏天", imgs, null, null);
        assertTrue(out.contains("至少需要 3 张"), "should demand >=3 images; got:\n" + out);
        assertFalse(out.contains("在线预览"), "must not produce a preview when refused");
    }
@ -54,7 +57,7 @@ class XhsPackageTest {
    @DisplayName("3 images → packaged; preview is image-first (images before the copy)")
    void packagesThreeImagesImageFirst() {
        String imgs = putImg("cover.png") + "," + putImg("c1.png") + "," + putImg("c2.png");
-        String out = tool.xhs_package("3天2夜厦门citywalk", "第一天去了鼓浪屿\n人不多", "厦门,citywalk,旅行", imgs, null);
+        String out = tool.xhs_package("3天2夜厦门citywalk", "第一天去了鼓浪屿\n人不多", "厦门,citywalk,旅行", imgs, null, null);

        assertTrue(out.contains("在线预览"), "should return a preview link");
        assertTrue(out.contains("素材下载"), "should return a material zip");
@ -76,7 +79,7 @@ class XhsPackageTest {
        // First ref uses the filename (id pattern can't parse it); two more are valid.
        String imgs = "/api/v1/files/generated/cover_xhs.png,"
                + putImg("c1.png") + "," + putImg("c2.png");
-        String out = tool.xhs_package("标题", "正文", "标签", imgs, null);
+        String out = tool.xhs_package("标题", "正文", "标签", imgs, null, null);

        assertTrue(out.contains("在线预览"), "name-based ref should self-heal to reach >=3; got:\n" + out);
        assertTrue(out.contains("3 张图"), "healed image should be counted");
--- a/mateclaw-ui/src/api/index.ts
+++ b/mateclaw-ui/src/api/index.ts
@ -187,6 +187,13 @@ export const chatApi = {
 }

 // ==================== Conversation ====================
+// Content calendar (read-only) — produced 公众号 / 小红书 pieces + lifecycle status.
+export const contentItemApi = {
+  list: (params?: { page?: number; size?: number; platform?: string; status?: string }) =>
+    http.get('/content-items', { params }),
+  summary: () => http.get('/content-items/summary'),
+}
+
 export const conversationApi = {
  list: () => http.get('/conversations'),
  /**
--- a/mateclaw-ui/src/i18n/locales/en-US.ts
+++ b/mateclaw-ui/src/i18n/locales/en-US.ts
@ -495,6 +495,27 @@ export default {
      failed: 'Failed',
    },
  },
+  contentCalendar: {
+    title: 'Content Calendar',
+    subtitle: 'Ledger of produced 公众号 / 小红书 pieces — what is packaged, published, or pending upload (auto-recorded on delivery).',
+    total: 'All',
+    packaged: 'Packaged',
+    draft: 'Draft',
+    published: 'Published',
+    allPlatforms: 'All platforms',
+    refresh: 'Refresh',
+    platform: 'Platform',
+    colTitle: 'Title',
+    topic: 'Topic',
+    status: 'Status',
+    createTime: 'Created',
+    publishTime: 'Published',
+    st_draft: 'Draft',
+    st_packaged: 'Packaged',
+    st_published: 'Published',
+    st_failed: 'Failed',
+    st_unknown: 'Unknown',
+  },
  nav: {
    dashboard: 'Dashboard',
    chat: 'Chat',
@ -505,6 +526,7 @@ export default {
    workspace: 'Workspace',
    agentContext: 'Agent Context',
    skills: 'Skills',
+    contentCalendar: 'Content Calendar',
    wiki: 'Wiki',
    enterprise: 'Enterprise',
    tools: 'Tools',
--- a/mateclaw-ui/src/i18n/locales/zh-CN.ts
+++ b/mateclaw-ui/src/i18n/locales/zh-CN.ts
@ -495,6 +495,27 @@ export default {
      failed: '失败',
    },
  },
+  contentCalendar: {
+    title: '内容日历',
+    subtitle: '公众号 / 小红书 产出台账 —— 哪些已打包、已发布、待上传，一目了然（交付时自动登记）。',
+    total: '全部',
+    packaged: '已打包',
+    draft: '草稿',
+    published: '已发布',
+    allPlatforms: '全部平台',
+    refresh: '刷新',
+    platform: '平台',
+    colTitle: '标题',
+    topic: '选题',
+    status: '状态',
+    createTime: '创建时间',
+    publishTime: '发布时间',
+    st_draft: '草稿',
+    st_packaged: '已打包',
+    st_published: '已发布',
+    st_failed: '失败',
+    st_unknown: '未知',
+  },
  nav: {
    dashboard: '仪表盘',
    chat: '对话',
@ -508,6 +529,7 @@ export default {
    workspace: '工作区',
    agentContext: '智能体上下文',
    skills: '技能',
+    contentCalendar: '内容日历',
    wiki: '知识库',
    enterprise: '企业场景',
    tools: '工具',
--- a/mateclaw-ui/src/router/index.ts
+++ b/mateclaw-ui/src/router/index.ts
@ -96,6 +96,12 @@ const router = createRouter({
          component: () => import('@/views/SkillMarket.vue'),
          meta: { title: 'Skills', requiredCapability: 'manage:skills' },
        },
+        {
+          path: 'content-calendar',
+          name: 'ContentCalendar',
+          component: () => import('@/views/ContentCalendar.vue'),
+          meta: { title: 'Content Calendar', requiredCapability: 'manage:agents' },
+        },
        // Tools 顶层入口已降级到 Settings ▸ Tools (Catalog) (RFC-090 Phase 1)
        // 旧路径 /tools 由下方 redirect 兼容
        {
--- a/mateclaw-ui/src/types/components.d.ts
+++ b/mateclaw-ui/src/types/components.d.ts
@ -23,9 +23,13 @@ declare module 'vue' {
    ElIcon: typeof import('element-plus/es/components/icon/index')['ElIcon']
    ElImageViewer: typeof import('element-plus/es/components/image-viewer/index')['ElImageViewer']
    ElOption: typeof import('element-plus/es/components/select/index')['ElOption']
+    ElPagination: typeof import('element-plus/es/components/pagination/index')['ElPagination']
    ElPopover: typeof import('element-plus/es/components/popover/index')['ElPopover']
    ElSelect: typeof import('element-plus/es/components/select/index')['ElSelect']
    ElSkeleton: typeof import('element-plus/es/components/skeleton/index')['ElSkeleton']
+    ElTable: typeof import('element-plus/es/components/table/index')['ElTable']
+    ElTableColumn: typeof import('element-plus/es/components/table/index')['ElTableColumn']
+    ElTag: typeof import('element-plus/es/components/tag/index')['ElTag']
    ElTooltip: typeof import('element-plus/es/components/tooltip/index')['ElTooltip']
    RouterLink: typeof import('vue-router')['RouterLink']
    RouterView: typeof import('vue-router')['RouterView']
--- a/mateclaw-ui/src/views/ContentCalendar.vue
+++ b/mateclaw-ui/src/views/ContentCalendar.vue
@ -0,0 +1,191 @@
+<template>
+  <div class="content-calendar">
+    <div class="page-header">
+      <h2>{{ t('contentCalendar.title') }}</h2>
+      <p class="subtitle">{{ t('contentCalendar.subtitle') }}</p>
+    </div>
+
+    <!-- Status summary cards -->
+    <div class="summary-cards">
+      <div
+        v-for="card in summaryCards"
+        :key="card.key"
+        class="summary-card"
+        :class="{ active: statusFilter === card.key }"
+        @click="toggleStatus(card.key)"
+      >
+        <div class="card-count">{{ card.count }}</div>
+        <div class="card-label">{{ card.label }}</div>
+      </div>
+    </div>
+
+    <!-- Filters -->
+    <div class="filters">
+      <el-select v-model="platformFilter" :placeholder="t('contentCalendar.allPlatforms')" clearable @change="reload">
+        <el-option label="公众号" value="gzh" />
+        <el-option label="小红书" value="xhs" />
+      </el-select>
+      <el-button :icon="RefreshIcon" @click="reload">{{ t('contentCalendar.refresh') }}</el-button>
+    </div>
+
+    <!-- Table -->
+    <el-table :data="rows" v-loading="loading" stripe class="calendar-table">
+      <el-table-column :label="t('contentCalendar.platform')" width="100">
+        <template #default="{ row }">
+          <el-tag size="small" :type="row.platform === 'gzh' ? 'success' : 'danger'">
+            {{ row.platform === 'gzh' ? '公众号' : '小红书' }}
+          </el-tag>
+        </template>
+      </el-table-column>
+      <el-table-column :label="t('contentCalendar.colTitle')" min-width="220">
+        <template #default="{ row }">
+          <a v-if="row.previewUrl" :href="row.previewUrl" target="_blank" rel="noopener" class="title-link">
+            {{ row.title || '(无标题)' }}
+          </a>
+          <span v-else>{{ row.title || '(无标题)' }}</span>
+        </template>
+      </el-table-column>
+      <el-table-column prop="topic" :label="t('contentCalendar.topic')" min-width="140" show-overflow-tooltip />
+      <el-table-column :label="t('contentCalendar.status')" width="120">
+        <template #default="{ row }">
+          <el-tag size="small" :type="statusTagType(row.status)">{{ statusLabel(row.status) }}</el-tag>
+        </template>
+      </el-table-column>
+      <el-table-column :label="t('contentCalendar.createTime')" width="170">
+        <template #default="{ row }">{{ formatTime(row.createTime) }}</template>
+      </el-table-column>
+      <el-table-column :label="t('contentCalendar.publishTime')" width="170">
+        <template #default="{ row }">{{ formatTime(row.publishTime) || '—' }}</template>
+      </el-table-column>
+    </el-table>
+
+    <div class="pager">
+      <el-pagination
+        layout="prev, pager, next, total"
+        :total="total"
+        :page-size="pageSize"
+        :current-page="page"
+        @current-change="onPageChange"
+      />
+    </div>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, computed, onMounted, h } from 'vue'
+import { useI18n } from 'vue-i18n'
+import { contentItemApi } from '@/api'
+
+const { t } = useI18n()
+
+// Minimal inline refresh icon (avoids an extra import).
+const RefreshIcon = () =>
+  h('svg', { width: 16, height: 16, viewBox: '0 0 24 24', fill: 'none', stroke: 'currentColor', 'stroke-width': 2 }, [
+    h('polyline', { points: '23 4 23 10 17 10' }),
+    h('path', { d: 'M20.49 15a9 9 0 1 1-2.12-9.36L23 10' }),
+  ])
+
+interface ContentItem {
+  id: string | number
+  platform: string
+  topic?: string
+  title?: string
+  status?: string
+  previewUrl?: string
+  createTime?: string
+  publishTime?: string
+}
+
+const rows = ref<ContentItem[]>([])
+const total = ref(0)
+const page = ref(1)
+const pageSize = ref(20)
+const loading = ref(false)
+const platformFilter = ref('')
+const statusFilter = ref('')
+const summary = ref<Record<string, number>>({})
+
+const summaryCards = computed(() => [
+  { key: '', label: t('contentCalendar.total'), count: summary.value.total ?? 0 },
+  { key: 'packaged', label: t('contentCalendar.packaged'), count: summary.value.packaged ?? 0 },
+  { key: 'draft', label: t('contentCalendar.draft'), count: summary.value.draft ?? 0 },
+  { key: 'published', label: t('contentCalendar.published'), count: summary.value.published ?? 0 },
+])
+
+function statusLabel(s?: string) {
+  return t(`contentCalendar.st_${s || 'unknown'}`)
+}
+function statusTagType(s?: string) {
+  return s === 'published' ? 'success' : s === 'draft' ? 'info' : s === 'failed' ? 'danger' : 'warning'
+}
+function formatTime(v?: string) {
+  if (!v) return ''
+  return v.replace('T', ' ').slice(0, 16)
+}
+
+async function loadSummary() {
+  try {
+    const res: any = await contentItemApi.summary()
+    summary.value = res.data || res || {}
+  } catch {
+    summary.value = {}
+  }
+}
+
+async function reload() {
+  loading.value = true
+  try {
+    const res: any = await contentItemApi.list({
+      page: page.value,
+      size: pageSize.value,
+      platform: platformFilter.value || undefined,
+      status: statusFilter.value || undefined,
+    })
+    const data = res.data || res
+    rows.value = data.records || data.list || []
+    total.value = Number(data.total ?? rows.value.length)
+  } catch {
+    // Degrade gracefully (e.g. backend not yet restarted with the new endpoint).
+    rows.value = []
+    total.value = 0
+  } finally {
+    loading.value = false
+  }
+}
+
+function toggleStatus(key: string) {
+  statusFilter.value = statusFilter.value === key ? '' : key
+  page.value = 1
+  reload()
+}
+function onPageChange(p: number) {
+  page.value = p
+  reload()
+}
+
+onMounted(() => {
+  loadSummary()
+  reload()
+})
+</script>
+
+<style scoped>
+.content-calendar { padding: 20px; max-width: 1200px; margin: 0 auto; }
+.page-header h2 { margin: 0 0 4px; font-size: 20px; }
+.subtitle { margin: 0 0 16px; color: var(--el-text-color-secondary); font-size: 13px; }
+.summary-cards { display: flex; gap: 12px; margin-bottom: 16px; flex-wrap: wrap; }
+.summary-card {
+  flex: 1; min-width: 120px; padding: 14px 16px; border-radius: 10px;
+  background: var(--el-fill-color-light); cursor: pointer; transition: all .15s;
+  border: 1px solid transparent;
+}
+.summary-card:hover { background: var(--el-fill-color); }
+.summary-card.active { border-color: var(--el-color-primary); }
+.card-count { font-size: 24px; font-weight: 700; }
+.card-label { font-size: 13px; color: var(--el-text-color-secondary); margin-top: 2px; }
+.filters { display: flex; gap: 10px; margin-bottom: 12px; }
+.calendar-table { width: 100%; }
+.title-link { color: var(--el-color-primary); text-decoration: none; }
+.title-link:hover { text-decoration: underline; }
+.pager { margin-top: 14px; display: flex; justify-content: flex-end; }
+</style>
--- a/mateclaw-ui/src/views/layout/MainLayout.vue
+++ b/mateclaw-ui/src/views/layout/MainLayout.vue
@ -510,6 +510,12 @@ const navGroups = computed(() => [
        icon: `<svg width="18" height="18" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polygon points="12 2 15.09 8.26 22 9.27 17 14.14 18.18 21.02 12 17.77 5.82 21.02 7 14.14 2 9.27 8.91 8.26 12 2"/></svg>`,
        requiredCapability: 'manage:skills',
      },
+      {
+        path: '/content-calendar',
+        label: t('nav.contentCalendar'),
+        icon: `<svg width="18" height="18" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><rect x="3" y="4" width="18" height="18" rx="2" ry="2"/><line x1="16" y1="2" x2="16" y2="6"/><line x1="8" y1="2" x2="8" y2="6"/><line x1="3" y1="10" x2="21" y2="10"/></svg>`,
+        requiredCapability: 'manage:agents',
+      },
      {
        path: '/plugins',
        label: t('nav.plugins'),