菜单
Publish a generated Agent-Media UGC Video to the user's connected TikTok, Instagram, or X. Connect channels (OAuth) and post or schedule via the REST API. Use after producing a video with make_ugc.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/gitroomhq/agent-media --skill publish-to-social命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Playbook for Agent-Media UGC Video — the one tool for UGC video on agent-media. Always call the single make_ugc skill: give it a `script` (any length) and optionally a person/image/character; it returns the finished captioned vertical video. Short script → one clip, long monologue → full multi-take (never trimmed), `broll_url` → narrated overlay. You never pick a sub-skill.
Two saved characters recording a podcast in ONE room — the camera cuts to whoever is speaking, and each actor stays in their IDENTICAL seat, desk and mic position across every cut. Provide character_a and character_b (saved char_… ids from list_characters, or https image URLs) and an ordered `script` of A/B dialogue turns (each { speaker: "A" | "B", line: "…" }). The pipeline renders ONE shared two-shot, locks a close-up per actor, animates every turn with native lip-synced Seedance voice (each actor keeps a consistent look AND voice across the whole episode), and hard-cuts the turns together as a 9:16 vertical video. Long turns auto-split into ≤15s takes. Captions are OPT-IN — ask the user first, then set subtitles:true.
The ONE tool for UGC video. Give a `script` (any length) and optionally a `person` description, an `image` (photo), or a `character` (saved char_… or sheet URL); it returns the finished vertical video. Short script → one clip; long monologue → full multi-take (never trimmed); pass `broll_url` → narrated b-roll overlay. Captions are OPT-IN — ASK the user if they want them (and which style) before generating; set `captions:true` only if they say yes. You never pick a sub-tool.
基于 SOC 职业分类
正在显示 SKILL.md
| name | Publish to Social |
| description | Publish a generated Agent-Media UGC Video to the user's connected TikTok, Instagram, or X. Connect channels (OAuth) and post or schedule via the REST API. Use after producing a video with make_ugc. |
| x-skill-slug | publish-to-social |
| x-skill-version | 1.0.0 |
Post a finished agent-media video (an R2 video_url from any of the video skills) to the user's connected social channels — TikTok, Instagram, or X. Works from three surfaces: REST, the CLI (agent-media social ...), and MCP tools (social_channels / social_connect / social_publish). All calls use the user's Authorization: Bearer ma_... token against https://api.agent-media.ai.
GET /v1/social/providers -> connectable networks: tiktok, instagram, instagram-standalone, x
POST /v1/social/connect { provider } -> { url } # the user opens this OAuth url and authorizes
GET /v1/social/channels -> { channels: [{ id, name, provider, profile, picture }] }
DELETE /v1/social/channels/:channelId -> disconnect
The connect step returns an OAuth URL the human must open and authorize — an agent cannot complete OAuth itself. Once authorized, the channel appears in /v1/social/channels with an id you pass to publish.
CLI: agent-media social providers · agent-media social connect x · agent-media social channels.
MCP: social_connect { provider } (returns the URL for the user) · social_channels.
curl -X POST https://api.agent-media.ai/v1/social/publish \
-H "Authorization: Bearer ma_..." -H "Content-Type: application/json" \
-d '{ "video_url": "https://pub-...r2.dev/generation-outputs/<user>/<job>/...mp4",
"channel_ids": ["<channel-id-from-/channels>"],
"caption": "made with agent-media",
"type": "now" }' # or "type":"schedule" + "date":"2026-06-01T10:00:00.000Z"
CLI: agent-media social publish --video <url> --channels <id,id> --caption "..." (add --at <iso> to schedule).
MCP: social_publish { video_url, channel_ids, caption, type }.
video_url must be an agent-media R2 URL (the output of a video skill) — agent-media re-hosts it on the publishing provider for you. channel_ids come from /v1/social/channels. Per-network requirements (e.g. X reply settings) are filled in server-side; you don't send them.
Returns { success: true, media_id, post_ids: ["..."] }. A real post was created only when post_ids is non-empty — treat an empty post_ids as a failure, not a success.
make_ugc → final_output.video_url.video_url to the chosen channel_ids; confirm post_ids came back.Note: social operations run on a shared rate budget — don't poll; connect once and publish on demand.