@remixmate/cli 0.9.70 → 0.9.72

package/README.md

@@ -187,6 +187,7 @@ This project contains 13 AI media generation skills covering the full pipeline f
 │  web-screenshot    Web page screenshot (png / jpg)          │
 │  web-record        Web page recording / storyboard video    │
 │  web-read          Web page text extraction (markdown/json) │
+│  gen-post          Post copy: Douyin / XHS / TikTok         │
 └─────────────────────────────────────────────────────────────┘
 ```
 
@@ -207,6 +208,7 @@ This project contains 13 AI media generation skills covering the full pipeline f
 │   ├── render-video/           # orchestration: Phase 3 Remotion render (canonical render_video.py)
 │   ├── export-jianying/        # orchestration: export Jianying draft ZIP
 │   ├── video-parser/           # tool: video deconstruction & analysis
+│   ├── gen-post/               # tool: social post copy per platform (http handler)
 │   ├── web-screenshot/         # tool: headless-browser screenshot (also hosts record.py)
 │   ├── web-record/             # tool: headless-browser recording (entry -> web-screenshot/scripts/record.py)
 │   └── web-read/               # tool: headless-browser text extraction (entry -> web-screenshot/scripts/read_page.py)
@@ -227,6 +229,7 @@ This project contains 13 AI media generation skills covering the full pipeline f
 | render-video | orchestration | job_id → Remotion render → upload | python `scripts/render_video.py` |
 | export-jianying | orchestration | Asset URLs → Jianying draft ZIP (auto-converts from RenderPlan) | python `scripts/gen_jianying_draft.py` |
 | video-parser | tool | Video → audio + ASR + keyframes + scene segmentation | python `scripts/parse_via_render.py` |
+| gen-post | tool | Post text for a finished video — title / description / hashtags for Douyin, Xiaohongshu, TikTok | http handler |
 | web-screenshot | tool | Web page screenshot (png / jpg) | python `scripts/screenshot.py` |
 | web-record | tool | Web page recording / scroll-through / storyboard video (webm -> mp4 -> VOD) | python `../web-screenshot/scripts/record.py` |
 | web-read | tool | Web page → main text (markdown / plain text / structured JSON) | python `../web-screenshot/scripts/read_page.py` |

package/README.zh-CN.md

@@ -141,6 +141,7 @@ Python 技能（13 个中的 9 个）需要 `python3 >= 3.10`。`web-screenshot`
 │  web-screenshot    网页截图（png / jpg）                     │
 │  web-record        网页录屏 / 滚动录屏 / 分镜视频            │
 │  web-read          网页正文抽取（markdown / 纯文本 / JSON）  │
+│  gen-post          发布文案（抖音 / 小红书 / TikTok）        │
 └─────────────────────────────────────────────────────────────┘
 ```
 
@@ -161,6 +162,7 @@ Python 技能（13 个中的 9 个）需要 `python3 >= 3.10`。`web-screenshot`
 │   ├── render-video/           # 编排: Phase 3 Remotion 渲染（含 render_video.py 实现）
 │   ├── export-jianying/        # 编排: 导出剪映草稿 ZIP
 │   ├── video-parser/           # 工具: 视频解构与分析
+│   ├── gen-post/               # 工具: 按平台生成发布文案（http handler）
 │   ├── web-screenshot/         # 工具: 无头浏览器截图（record.py 也放在这里）
 │   ├── web-record/             # 工具: 无头浏览器录屏（入口指向 web-screenshot/scripts/record.py）
 │   └── web-read/               # 工具: 无头浏览器正文抽取（入口指向 web-screenshot/scripts/read_page.py）
@@ -181,6 +183,7 @@ Python 技能（13 个中的 9 个）需要 `python3 >= 3.10`。`web-screenshot`
 | render-video | 编排 | job_id → Remotion 渲染 → 上传 | python `scripts/render_video.py` |
 | export-jianying | 编排 | 素材 URL → 剪映草稿 ZIP（支持从 RenderPlan 自动转换） | python `scripts/gen_jianying_draft.py` |
 | video-parser | 工具 | 视频 → 音频 + ASR + 关键帧 + 场景分段 | python `scripts/parse_via_render.py` |
+| gen-post | 工具 | 成片的发布文案：抖音 / 小红书 / TikTok 的标题、描述、话题标签 | http handler |
 | web-screenshot | 工具 | 网页截图（png / jpg） | python `scripts/screenshot.py` |
 | web-record | 工具 | 网页录屏 / 滚动录屏 / 分镜视频（webm → mp4 → VOD） | python `../web-screenshot/scripts/record.py` |
 | web-read | 工具 | 网页正文抽取（markdown / 纯文本 / 结构化 JSON） | python `../web-screenshot/scripts/read_page.py` |

package/dist/handlers/gen-post.d.ts

@@ -0,0 +1,22 @@
+/**
+ * gen-post handler — POST ab-api /publish/caption/generate.
+ *
+ * Generates the text of a social post — per-platform title / description / tags —
+ * for a finished video (the video itself is not touched or uploaded): the same endpoint and prompts the web studio's 「发布文案」
+ * panel uses. Prompt wording, compliance rules and the platform roster all live
+ * in ab-api (publish_prompts.go + the user's saved prompt overrides); this
+ * handler only collects context and forwards it.
+ *
+ * Behavior:
+ *   - --platform is repeatable and/or comma-separated; `all` (the default) expands
+ *     to the three platforms the web panel shows. Each platform is one backend
+ *     call (one LLM run, billed separately), issued concurrently.
+ *   - --render-id links the result to a studio take so it shows up in the web
+ *     panel; without it the backend generates without persisting anything.
+ *   - --instruction is the user's "how should it read" request. With a draft
+ *     (--title / --description / --tag) it rewrites that caption — single platform
+ *     only, since the draft belongs to one platform. Without a draft it is an extra
+ *     requirement applied to a fresh generation on every requested platform.
+ */
+import type { HandlerContext, HandlerInput } from './index.js';
+export declare function genPost(input: HandlerInput, ctxIn: HandlerContext): Promise<void>;

package/dist/handlers/gen-post.js

@@ -0,0 +1,192 @@
+/**
+ * gen-post handler — POST ab-api /publish/caption/generate.
+ *
+ * Generates the text of a social post — per-platform title / description / tags —
+ * for a finished video (the video itself is not touched or uploaded): the same endpoint and prompts the web studio's 「发布文案」
+ * panel uses. Prompt wording, compliance rules and the platform roster all live
+ * in ab-api (publish_prompts.go + the user's saved prompt overrides); this
+ * handler only collects context and forwards it.
+ *
+ * Behavior:
+ *   - --platform is repeatable and/or comma-separated; `all` (the default) expands
+ *     to the three platforms the web panel shows. Each platform is one backend
+ *     call (one LLM run, billed separately), issued concurrently.
+ *   - --render-id links the result to a studio take so it shows up in the web
+ *     panel; without it the backend generates without persisting anything.
+ *   - --instruction is the user's "how should it read" request. With a draft
+ *     (--title / --description / --tag) it rewrites that caption — single platform
+ *     only, since the draft belongs to one platform. Without a draft it is an extra
+ *     requirement applied to a fresh generation on every requested platform.
+ */
+import { EXIT, mmPost, resolveHttpContext, SkillError } from '../http.js';
+import { emitProgress } from '../progress.js';
+import { isTrue, toNumber, toStringArray } from './shared.js';
+/** Platforms the web panel enables by default; `all` expands to these. */
+const DEFAULT_PLATFORMS = ['douyin', 'xiaohongshu', 'tiktok'];
+/** Friendly aliases → backend platform codes. Unknown values pass through for ab-api to validate. */
+const PLATFORM_ALIASES = {
+    dy: 'douyin',
+    抖音: 'douyin',
+    xhs: 'xiaohongshu',
+    rednote: 'xiaohongshu',
+    小红书: 'xiaohongshu',
+    tt: 'tiktok',
+};
+/** Display names for human-readable output only. */
+const PLATFORM_LABELS = {
+    douyin: '抖音 Douyin',
+    xiaohongshu: '小红书 Xiaohongshu',
+    tiktok: 'TikTok',
+};
+/** Backend waits on a non-streaming LLM call (80s server-side cap); leave headroom. */
+const GENERATE_TIMEOUT_MS = 90_000;
+function resolvePlatforms(value) {
+    const raw = toStringArray(value)
+        .flatMap((v) => v.split(/[,，\s]+/))
+        .map((v) => v.trim())
+        .filter(Boolean);
+    if (raw.length === 0)
+        return [...DEFAULT_PLATFORMS];
+    const out = [];
+    for (const item of raw) {
+        const key = item.toLowerCase();
+        // 「全部」: the web param card writes option labels into the message, and the agent
+        // may forward the label verbatim.
+        const codes = key === 'all' || item === '全部' ? [...DEFAULT_PLATFORMS] : [PLATFORM_ALIASES[item] ?? PLATFORM_ALIASES[key] ?? key];
+        for (const code of codes)
+            if (!out.includes(code))
+                out.push(code);
+    }
+    return out;
+}
+function optString(value) {
+    if (typeof value !== 'string')
+        return undefined;
+    const trimmed = value.trim();
+    return trimmed || undefined;
+}
+/** Tags may arrive as `--tag a --tag b`, `--tag "a,b"`, or with a leading `#`. */
+function normalizeTags(value) {
+    return toStringArray(value)
+        .flatMap((v) => v.split(/[,，]+/))
+        .map((v) => v.trim().replace(/^#+/, ''))
+        .filter(Boolean);
+}
+async function generateOne(ctx, platform, base) {
+    emitProgress({ phase: 'gen-post:request', platform });
+    try {
+        const caption = await mmPost(ctx, '/publish/caption/generate', { ...base, platform }, { timeoutMs: GENERATE_TIMEOUT_MS });
+        if (!caption || (!caption.title && !caption.description)) {
+            return { platform, error: 'backend returned an empty caption' };
+        }
+        emitProgress({ phase: 'gen-post:done', platform });
+        return { platform, caption: { ...caption, platform, tags: caption.tags ?? [] } };
+    }
+    catch (err) {
+        // Auth failures abort the whole run: every other platform would fail the same way.
+        if (err instanceof SkillError && err.exitCode === EXIT.NOT_AUTHENTICATED)
+            throw err;
+        return { platform, error: err instanceof Error ? err.message : String(err) };
+    }
+}
+function printHuman(results) {
+    for (const r of results) {
+        const label = PLATFORM_LABELS[r.platform] ?? r.platform;
+        process.stdout.write(`\n━━━━━━━━ ${label} ━━━━━━━━\n`);
+        if (!r.caption) {
+            process.stdout.write(`❌ ${r.error ?? 'generation failed'}\n`);
+            continue;
+        }
+        const { title, description, tags } = r.caption;
+        process.stdout.write(`标题 / Title:\n${title}\n\n`);
+        process.stdout.write(`描述 / Description:\n${description}\n\n`);
+        process.stdout.write(`标签 / Tags:\n${tags.map((t) => `#${t}`).join(' ')}\n`);
+    }
+    const ok = results.filter((r) => r.caption).length;
+    process.stdout.write(`\n🎉 Generated publish copy for ${ok}/${results.length} platform(s)\n`);
+}
+export async function genPost(input, ctxIn) {
+    const topic = optString(input.topic);
+    const hook = optString(input.hook);
+    const instruction = optString(input.instruction);
+    const renderId = optString(input.render_id);
+    const platforms = resolvePlatforms(input.platform);
+    const current = {
+        title: optString(input.title) ?? '',
+        description: optString(input.description) ?? '',
+        tags: normalizeTags(input.tag),
+    };
+    const hasDraft = !!(current.title || current.description || current.tags.length);
+    if (instruction && instruction.length > 500) {
+        throw new SkillError('❌ --instruction must be at most 500 characters');
+    }
+    // With a draft, --instruction rewrites that one platform's caption. Without one it is
+    // an extra requirement for a fresh generation, which needs the video's context.
+    const rewriting = !!instruction && hasDraft;
+    if (hasDraft && !instruction) {
+        throw new SkillError('❌ --title / --description / --tag are the draft to rewrite; pass --instruction with them');
+    }
+    if (rewriting && platforms.length !== 1) {
+        throw new SkillError('❌ rewriting a caption (--instruction with --title / --description / --tag) needs exactly one --platform');
+    }
+    if (!rewriting && !topic && !hook) {
+        throw new SkillError('❌ Please provide --topic (what the video is about); --hook is optional extra context');
+    }
+    const durationSec = toNumber(input.duration, 'duration');
+    if (durationSec !== undefined && durationSec < 0) {
+        throw new SkillError('❌ --duration must be a non-negative number of seconds');
+    }
+    const base = {};
+    if (renderId)
+        base.renderId = renderId;
+    const videoProjectId = optString(input.video_project_id);
+    if (videoProjectId)
+        base.videoProjectId = videoProjectId;
+    if (topic)
+        base.topic = topic;
+    if (hook)
+        base.hook = hook;
+    const sourceUrl = optString(input.source_url);
+    if (sourceUrl)
+        base.sourceUrl = sourceUrl;
+    if (durationSec !== undefined)
+        base.durationSec = durationSec;
+    const aspectRatio = optString(input.aspect_ratio);
+    if (aspectRatio)
+        base.aspectRatio = aspectRatio;
+    const language = optString(input.language);
+    if (language)
+        base.language = language;
+    const promptOverride = optString(input.prompt);
+    if (promptOverride)
+        base.promptOverride = promptOverride;
+    if (instruction)
+        base.instruction = instruction;
+    if (rewriting)
+        base.current = current;
+    const ctx = await resolveHttpContext(ctxIn.skillName, {
+        apiBaseUrl: input.api_base_url,
+        privateToken: input.priv_token,
+        preflightToken: ctxIn.auth?.token,
+    });
+    const results = await Promise.all(platforms.map((p) => generateOne(ctx, p, base)));
+    const succeeded = results.filter((r) => r.caption);
+    if (isTrue(input.json_output)) {
+        process.stdout.write(JSON.stringify({
+            captions: succeeded.map((r) => ({
+                platform: r.platform,
+                title: r.caption.title,
+                description: r.caption.description,
+                tags: r.caption.tags,
+            })),
+            errors: results.filter((r) => !r.caption).map((r) => ({ platform: r.platform, error: r.error })),
+            persisted: !!renderId,
+        }) + '\n');
+    }
+    else {
+        printHuman(results);
+    }
+    if (succeeded.length === 0) {
+        throw new SkillError(`❌ publish copy generation failed for every platform: ${results.map((r) => `${r.platform}: ${r.error}`).join('; ')}`);
+    }
+}

package/dist/handlers/index.js

@@ -9,10 +9,12 @@ import { genVoice } from './gen-voice.js';
 import { genVideo } from './gen-video.js';
 import { genDigitalHuman } from './gen-digital-human.js';
 import { templateInit } from './template-init.js';
+import { genPost } from './gen-post.js';
 export const HANDLERS = {
     'gen-image': genImage,
     'gen-voice': genVoice,
     'gen-video': genVideo,
     'gen-digital-human': genDigitalHuman,
     'template-init': templateInit,
+    'gen-post': genPost,
 };

package/dist/manifest.json

@@ -1,7 +1,7 @@
 {
   "schemaVersion": 1,
-  "version": "0.9.70",
-  "generatedAt": "2026-10-06T15:32:49.285Z",
+  "version": "0.9.72",
+  "generatedAt": "2026-10-08T04:30:08.225Z",
   "skills": [
     {
       "id": "doc-parse",
@@ -411,6 +411,123 @@
         ]
       }
     },
+    {
+      "id": "gen-post",
+      "toolName": "gen_post",
+      "tier": "tool",
+      "category": "consuming",
+      "title": "Social Post Copy",
+      "summary": "Write the text of a social post for a finished video — title, description and hashtags — per platform (text only; it does not upload or publish the video): Douyin (抖音), Xiaohongshu (小红书) and TikTok. Same prompts as the web studio's publish panel, including the platform's length limits and ad-law / community-guideline compliance rules, and the user's saved prompt overrides. Generates all three platforms by default; pass `platform` to pick. Fill the video context yourself from the conversation — topic, hook, source URL, duration, aspect ratio — rather than asking the user; it only writes from what it is given. `instruction` carries the user's own request: alone it shapes a fresh generation, and with the current title/description/tags (from an earlier run in the conversation) it rewrites that caption instead. Each platform is one LLM call and costs credits.",
+      "triggers": [
+        "Publish copy / caption / title + description + hashtags for a video",
+        "发布文案、视频标题、话题标签、抖音文案、小红书笔记文案、TikTok caption",
+        "Rewrite or polish an existing video caption for a platform"
+      ],
+      "entry": {
+        "type": "http",
+        "handler": "gen-post"
+      },
+      "runtime": "node",
+      "envVars": [
+        "PRIV_TOKEN",
+        "MM_API_BASE_URL",
+        "AGENT_NAME"
+      ],
+      "parameters": {
+        "type": "object",
+        "properties": {
+          "topic": {
+            "type": "string",
+            "description": "What the video is about — the main input for the copy. Take it from the conversation (video title, script, source page). Required unless rewriting an existing caption."
+          },
+          "platform": {
+            "type": "array",
+            "items": {
+              "type": "string"
+            },
+            "description": "Target platforms: douyin / xiaohongshu / tiktok, or all (default: all three). Repeatable or comma-separated."
+          },
+          "hook": {
+            "type": "string",
+            "description": "The video's opening hook line, if known"
+          },
+          "source_url": {
+            "type": "string",
+            "description": "Project / source URL the video is about, if any (used as context; never written into the copy as a link)"
+          },
+          "duration": {
+            "type": "number",
+            "description": "Video duration in seconds"
+          },
+          "aspect_ratio": {
+            "type": "string",
+            "description": "Video aspect ratio, e.g. 9:16 (default 9:16)"
+          },
+          "language": {
+            "type": "string",
+            "description": "Output language. Default: English for tiktok, 简体中文 for the others."
+          },
+          "instruction": {
+            "type": "string",
+            "description": "The user's request for how the copy should read, e.g. 'shorter title, drop the emoji'. Alone: applied as an extra requirement to a fresh generation. With the current title / description / tag: rewrites that caption (exactly one platform). Max 500 characters."
+          },
+          "title": {
+            "type": "string",
+            "description": "Current title to rewrite (only with `instruction`)"
+          },
+          "description": {
+            "type": "string",
+            "description": "Current description to rewrite (only with `instruction`)"
+          },
+          "tag": {
+            "type": "array",
+            "items": {
+              "type": "string"
+            },
+            "description": "Current tags to rewrite (only with `instruction`). Repeatable or comma-separated; a leading # is stripped."
+          },
+          "prompt": {
+            "type": "string",
+            "description": "Replace the platform prompt for this run only. Placeholders: {{topic}} {{hook}} {{sourceUrl}} {{durationSec}} {{aspectRatio}} {{language}}. The output must still be a JSON object with title / description / tags."
+          },
+          "render_id": {
+            "type": "string",
+            "description": "Studio render (take) id. When given, the result is saved to that take and shows up in the web publish panel; otherwise nothing is saved."
+          },
+          "video_project_id": {
+            "type": "string",
+            "description": "Studio video project id of the take (only with `render_id`)"
+          },
+          "json_output": {
+            "type": "boolean",
+            "description": "Emit a JSON result ({ captions: [{ platform, title, description, tags }], errors, persisted }) instead of human-readable output"
+          }
+        },
+        "required": []
+      },
+      "ui": {
+        "primary": [
+          "platform",
+          "instruction"
+        ],
+        "advanced": [],
+        "hidden": [
+          "topic",
+          "hook",
+          "source_url",
+          "duration",
+          "aspect_ratio",
+          "language",
+          "title",
+          "description",
+          "tag",
+          "prompt",
+          "render_id",
+          "video_project_id",
+          "json_output"
+        ]
+      }
+    },
     {
       "id": "gen-script",
       "toolName": "gen_script",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@remixmate/cli",
-  "version": "0.9.70",
+  "version": "0.9.72",
   "description": "AI media generation skills for Claude Code / Codex — 12 skills covering image, video, voice, digital human, web screenshot, web recording, script, template registry, rendering, Jianying export, and video deconstruction.",
   "type": "module",
   "main": "dist/index.js",

package/skills/gen-post/SKILL.md

@@ -0,0 +1,114 @@
+---
+name: gen-post
+description: |
+  Social-post skill: write the post text — title, description and hashtags — for a finished video, per platform: Douyin (抖音), Xiaohongshu (小红书) and TikTok. Text only; it does not upload or publish the video. Backed by ab-api's `/publish/caption/generate`, the same prompts the web studio's publish panel uses.
+
+  Use this skill immediately whenever the user asks for any of:
+  - Publish copy / caption / title + description + hashtags for a video
+  - 发布文案、视频标题、话题标签、抖音文案、小红书笔记文案、TikTok caption
+  - Rewrite or polish an existing video caption for a platform
+triggers:
+  - Publish copy / caption / title + description + hashtags for a video
+  - 发布文案、视频标题、话题标签、抖音文案、小红书笔记文案、TikTok caption
+  - Rewrite or polish an existing video caption for a platform
+---
+
+# Social Post Skill
+
+Wraps ab-api's `POST /publish/caption/generate`. Platform prompts (length limits, tone, and
+the ad-law / community-guideline compliance rules) live in the backend, and a user's saved
+prompt overrides from the web panel apply here too.
+
+| `--platform` | Default language | Shape |
+|--------------|------------------|-------|
+| `douyin` (`dy`) | 简体中文 | short hook title, 60-150 字 description, 3-5 tags |
+| `xiaohongshu` (`xhs`) | 简体中文 | note-style title ≤ 20 字, 200-400 字 body, 5-8 tags |
+| `tiktok` (`tt`) | English | first-line title, short caption, 3-5 hashtags |
+
+Default is all three. Each platform is one LLM call, run concurrently, charged separately.
+
+## Where the inputs come from
+
+**You fill the video context; the user only says how they want it to read.** Take
+`--topic` / `--hook` / `--source-url` / `--duration` / `--aspect-ratio` from the conversation —
+the video's title, the script, the page it was made from, the render result. Don't ask the
+user for them. The copy is written only from what you pass — anything missing is treated as
+"not provided", never invented — so pass everything you know.
+
+The user's own words ("标题去掉 emoji", "语气轻松一点", "只要小红书") go to `--instruction`
+(and `--platform` when they name platforms). If they gave no request, just generate.
+
+## Generate
+
+```bash
+remixmate gen-post \
+  --topic "<what the video is about>" \
+  --hook "<opening line>" \
+  --duration 45 --aspect-ratio 9:16
+```
+
+```bash
+# One platform, machine-readable
+remixmate gen-post --topic "<topic>" --platform xhs --json-output
+```
+
+```bash
+# The user's request shapes a fresh generation on every platform
+remixmate gen-post --topic "<topic>" --instruction "语气轻松一点，标题不要 emoji"
+```
+
+## Rewrite
+
+When a caption was already generated earlier in the conversation and the user asks to change
+it, pass that caption back as the draft: `--instruction` plus `--title` / `--description` /
+`--tag` rewrites it instead of starting over. One platform per call. Fields the instruction
+doesn't mention are kept. Compliance rules still win either way: asking for superlatives,
+guarantees or off-platform links gets a compliant rewording, not a literal one.
+
+```bash
+remixmate gen-post --platform douyin \
+  --instruction "标题去掉 emoji，再短一点" \
+  --title "<current title>" --description "<current description>" \
+  --tag tagA --tag tagB
+```
+
+## Flags
+
+| Flag | Description | Default |
+|------|-------------|---------|
+| `--topic` | What the video is about (required unless rewriting) | — |
+| `--platform` | `douyin` / `xiaohongshu` / `tiktok` / `all`; repeatable or comma-separated | `all` |
+| `--hook` | Opening hook line | none |
+| `--source-url` | Project/source URL, context only | none |
+| `--duration` | Seconds | none |
+| `--aspect-ratio` | e.g. `9:16` | `9:16` |
+| `--language` | Output language | per platform (see table) |
+| `--instruction` | The user's request, ≤ 500 chars; rewrites when a draft is given | none |
+| `--title` / `--description` / `--tag` | Current caption to rewrite | none |
+| `--prompt` | Replace the platform prompt for this run; placeholders `{{topic}}` `{{hook}}` `{{sourceUrl}}` `{{durationSec}}` `{{aspectRatio}}` `{{language}}` | saved/default prompt |
+| `--render-id` / `--video-project-id` | Save the result to a studio take (shows in the web publish panel) | not saved |
+| `--json-output` | `{ captions: [{ platform, title, description, tags }], errors, persisted }` | off |
+| `--api-base-url` / `--priv-token` | Override API root / token | env |
+
+Tags come back without `#`; the human-readable output adds it.
+
+## Output
+
+Show the user each platform's title, description and tags as-is so they can copy them. If
+some platforms failed, the rest still print; the run exits non-zero only when every platform
+failed.
+
+## Credits
+
+Each platform charges credits. Relay the footer when it appears:
+
+```
+💳 Charged <credits> credits · balance <balance>
+```
+
+## Error handling
+
+- **Token missing / 401**: set `PRIV_TOKEN` or run `remixmate login`.
+- **`不支持的平台`**: the platform code is not one the backend knows.
+- **`模型响应超时` / timeout**: retry; generating fewer platforms at once also helps.
+- **Rewrite errors**: a draft (`--title` / `--description` / `--tag`) needs `--instruction` and exactly one `--platform`.

package/skills/gen-post/skill.json

@@ -0,0 +1,112 @@
+{
+  "name": "gen-post",
+  "toolName": "gen_post",
+  "tier": "tool",
+  "category": "consuming",
+  "title": "Social Post Copy",
+  "description": "Write the text of a social post for a finished video — title, description and hashtags — per platform (text only; it does not upload or publish the video): Douyin (抖音), Xiaohongshu (小红书) and TikTok. Same prompts as the web studio's publish panel, including the platform's length limits and ad-law / community-guideline compliance rules, and the user's saved prompt overrides. Generates all three platforms by default; pass `platform` to pick. Fill the video context yourself from the conversation — topic, hook, source URL, duration, aspect ratio — rather than asking the user; it only writes from what it is given. `instruction` carries the user's own request: alone it shapes a fresh generation, and with the current title/description/tags (from an earlier run in the conversation) it rewrites that caption instead. Each platform is one LLM call and costs credits.",
+  "auth": "required",
+  "envVars": [
+    "PRIV_TOKEN",
+    "MM_API_BASE_URL",
+    "AGENT_NAME"
+  ],
+  "entry": {
+    "type": "http",
+    "handler": "gen-post"
+  },
+  "parameters": {
+    "type": "object",
+    "properties": {
+      "topic": {
+        "type": "string",
+        "description": "What the video is about — the main input for the copy. Take it from the conversation (video title, script, source page). Required unless rewriting an existing caption."
+      },
+      "platform": {
+        "type": "array",
+        "items": {
+          "type": "string"
+        },
+        "description": "Target platforms: douyin / xiaohongshu / tiktok, or all (default: all three). Repeatable or comma-separated."
+      },
+      "hook": {
+        "type": "string",
+        "description": "The video's opening hook line, if known"
+      },
+      "source_url": {
+        "type": "string",
+        "description": "Project / source URL the video is about, if any (used as context; never written into the copy as a link)"
+      },
+      "duration": {
+        "type": "number",
+        "description": "Video duration in seconds"
+      },
+      "aspect_ratio": {
+        "type": "string",
+        "description": "Video aspect ratio, e.g. 9:16 (default 9:16)"
+      },
+      "language": {
+        "type": "string",
+        "description": "Output language. Default: English for tiktok, 简体中文 for the others."
+      },
+      "instruction": {
+        "type": "string",
+        "description": "The user's request for how the copy should read, e.g. 'shorter title, drop the emoji'. Alone: applied as an extra requirement to a fresh generation. With the current title / description / tag: rewrites that caption (exactly one platform). Max 500 characters."
+      },
+      "title": {
+        "type": "string",
+        "description": "Current title to rewrite (only with `instruction`)"
+      },
+      "description": {
+        "type": "string",
+        "description": "Current description to rewrite (only with `instruction`)"
+      },
+      "tag": {
+        "type": "array",
+        "items": {
+          "type": "string"
+        },
+        "description": "Current tags to rewrite (only with `instruction`). Repeatable or comma-separated; a leading # is stripped."
+      },
+      "prompt": {
+        "type": "string",
+        "description": "Replace the platform prompt for this run only. Placeholders: {{topic}} {{hook}} {{sourceUrl}} {{durationSec}} {{aspectRatio}} {{language}}. The output must still be a JSON object with title / description / tags."
+      },
+      "render_id": {
+        "type": "string",
+        "description": "Studio render (take) id. When given, the result is saved to that take and shows up in the web publish panel; otherwise nothing is saved."
+      },
+      "video_project_id": {
+        "type": "string",
+        "description": "Studio video project id of the take (only with `render_id`)"
+      },
+      "json_output": {
+        "type": "boolean",
+        "description": "Emit a JSON result ({ captions: [{ platform, title, description, tags }], errors, persisted }) instead of human-readable output"
+      }
+    },
+    "required": []
+  },
+  "ui": {
+    "primary": [
+      "platform",
+      "instruction"
+    ],
+    "advanced": [],
+    "hidden": [
+      "topic",
+      "hook",
+      "source_url",
+      "duration",
+      "aspect_ratio",
+      "language",
+      "title",
+      "description",
+      "tag",
+      "prompt",
+      "render_id",
+      "video_project_id",
+      "json_output"
+    ]
+  }
+}

package/skills/gen-post/version.json

@@ -0,0 +1,7 @@
+{
+  "skillName": "gen-post",
+  "repoName": "agent-skill-media-maker",
+  "skillId": "pending-gen-post",
+  "version": "V1",
+  "skillDescription": "Social-post skill: write the post text — title, description and hashtags — for a finished video, per platform: Douyin (抖音), Xiaohongshu (小红书) and TikTok. Text only; it does not upload or publish the video. Backed by ab-api's `/publish/caption/generate` (the same prompts as the web studio's publish panel).\n\nUse this skill immediately whenever the user asks for any of:\n- Publish copy / caption / title + description + hashtags for a video\n- 发布文案、视频标题、话题标签、抖音文案、小红书笔记文案、TikTok caption\n- Rewrite or polish an existing video caption for a platform"
+}