AGENT SKILL

revid-blog-to-avatar-video

Turn a blog post URL into a talking-head avatar video — the avatar reads a summarized script of the post against a clean background. Use when the user wants a personal/expert delivery vs an edited promo. Calls the Revid MCP server (render_video → get_project_status).

License 许可
Apache-2.0
最近更新
2026-11-14
Stars
★ 1.6k
Forks
491

01 前言 Front Matter

name
revid-blog-to-avatar-video
metadata
{ "openclaw": { "requires": { "config": [ "REVID_API_KEY" ] } } }
description
Turn a blog post URL into a talking-head avatar video — the avatar reads a summarized script of the post against a clean background. Use when the user wants a personal/expert delivery vs an edited promo. Calls the Revid MCP server (render_video → get_project_status).

02 资源文件 2

03 SKILL.md 正文

Blog post → talking-head avatar video

Calls the Revid MCP server (https://www.revid.ai/api/mcp). Install once — see revid-api-foundations.

Take any blog/article URL and produce a vertical (or square) talking-head video with a chosen avatar reading a summarized version of the post.

When to use this skill

  • Source is a blog post / opinion piece / explainer with substantial body text.
  • Output should feel like a person delivering the take, not an edited promo with stock b-roll.
  • An avatar (image URL or characterId) is available, or the user accepts the default avatar.
  • For an edited short with stock visuals, use revid-article-to-short instead.

Inputs

Field Required Notes
url yes Blog post URL
avatar.url or characterIds[] yes The face. Either an image URL or a saved consistent character ID (see character mgmt).
aspectRatio no Default 9:16. Use 1:1 for LinkedIn.
voiceId no Match it to the avatar's tone if known.
targetDuration no Default 60 (s) — talking heads can run longer.

Step-by-step

  1. Validate the URL.
  2. If the user gave an avatar image URL, set avatar.url. If they gave a saved character ID, set characterIds: [id] (and leave avatar omitted).
  3. Call MCP tool render_video with the payload below — returns data.pid. Then poll MCP tool get_project_status with that pid every 5–8 s until data.status === "ready". Optionally call export_video for a freshly named mp4.
  4. Return videoUrl.

render_video arguments

{
  "workflow": "article-to-video",
  "source": {
    "url": "{BLOG_URL}",
    "scrapingPrompt": "Extract the article body. Skip header, navigation, related posts, and footer."
  },
  "aspectRatio": "9:16",
  "avatar": {
    "enabled": true,
    "url": "{AVATAR_IMAGE_URL}",
    "removeBackground": true,
    "imageModel": "good"
  },
  "voice": {
    "enabled": true,
    "voiceId": "aria-en-us",
    "stability": 0.65,
    "speed": 1.0,
    "language": "en-US",
    "enhanceAudio": true
  },
  "captions": { "enabled": true, "position": "bottom", "autoCrop": true },
  "music":    { "enabled": false },
  "media": {
    "type": "moving-image",
    "density": "low",
    "animation": "soft",
    "placeAvatarInContext": true
  },
  "options": {
    "targetDuration": 60,
    "summarizationPreference": "summarize",
    "hasToGenerateCover": true
  },
  "render": { "resolution": "1080p", "frameRate": 30 }
}

Notes:

  • placeAvatarInContext: true composites the avatar over a relevant background (vs a plain green-screen feel).
  • media.density: "low" keeps cuts minimal so the talking head can carry the video.
  • music.enabled: false is the default — voice-driven content reads better without competing audio.

Consistent characters

If the user wants the same face across many posts, create a character once and reuse the ID:

  1. Call create_character once:
    { "name": "Maya", "description": "Friendly host in her thirties, short dark hair", "imageUrl": "https://cdn.example.com/maya.jpg" }
    
    It returns the character id.
  2. Pass it in every render_video call:
    { "characterIds": ["<id>"], "avatar": { "enabled": true } }
    

List existing characters with list_characters.

Examples

  • examples/blog-to-avatar.json — copy-paste body for render_video (also a valid POST body for the direct HTTPS fallback).
  • examples/run.sh — bash smoke test using the direct HTTPS fallback (POST /api/public/v3/render → GET /status). Useful when you don't have an MCP client at hand.

Polling

Call get_project_status with the pid returned by render_video. Stop when data.status === "ready"; fail when data.status === "failed". Cadence: 5 s, then 8 s once progress > 30. Full pseudocode: revid-api-foundations § Polling.

Failure modes

Symptom Fix
Avatar lip-sync drifts on long copy Lower targetDuration to 45 s, or switch summarizationPreference: "summarize" (already on).
Avatar background bleeds into video Set avatar.removeBackground: true (default). For stubborn cases, pre-process the avatar image to a transparent PNG.
Background visuals distract from face media.density: "low" and media.animation: "soft" (already on). For pure plain background, set media.type: "custom" + media.useOnlyProvided: true with a single neutral asset.
Voice doesn't match the avatar Set voice.voiceId explicitly. The default voice is gendered female English — always override for other languages or personas.
scrape failed Same as in revid-article-to-short: pre-scrape the post and switch to script-to-video with the avatar block intact.
ok: false, error: "insufficient_credits" Drop media.quality to "standard" or render.resolution to "720p", or call buy_credit_pack.

See also

04 同仓库更多 Skills

hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-review-with-annotations

Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-find-and-synthesize-context

Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-update-docs

Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-records

Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-interactive-html

Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-collaborate-threads

Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.

★ 1.6k