blitzreels-faceless
10
总安装量
6
周安装量
#30976
全站排名
安装命令
npx skills add https://github.com/blitzreels/agent-skills --skill blitzreels-faceless
Agent 安装分布
amp
6
opencode
6
kimi-cli
6
codex
6
github-copilot
6
gemini-cli
6
Skill 文档
BlitzReels Faceless Video Generation
Generate AI-powered faceless videos from a topic or script.
Cost Warning: Faceless generation consumes AI credits (AI script writing, image generation, voiceover, video assembly). You MUST present a plan to the user and get approval before calling any expensive endpoint.
Quick Start
# Full pipeline: topic â video
bash scripts/faceless.sh --topic "5 productivity tips" --duration 30 --voice rachel --style cinematic
# Dry run â show plan without executing
bash scripts/faceless.sh --topic "5 productivity tips" --dry-run
# From a user-provided script (paragraph breaks = scene breaks)
bash scripts/faceless.sh --script "First scene narration.\n\nSecond scene narration." --duration 60
Required Workflow
Always use plan mode. Before generating, present this plan to the user for approval:
- Topic or script â What the video is about
- Duration â 10â120 seconds (drives scene count)
- Visual style â See
references/visual-styles.mdorGET /visual-styles - Voice â See
references/voices.mdorGET /voices - Image model â See
references/models.md(default: gemini-2.5-flash) - Video model â See
references/models.md(default: kling-2.1) - Output mode â Full video, storyboard-only (images), or images + audio
- Captions â Enabled by default; optionally pick a caption style
- Plan summary â Present all choices + “this will consume AI credits” warning
- User approval â Execute
Input Modes
| Mode | Flag | Behavior |
|---|---|---|
| Topic | --topic "..." |
AI writes the script based on the topic |
| Script | --script "..." |
User provides exact narration. Paragraph breaks (\n\n) become scene breaks. |
If both are provided, --script takes precedence.
Scripts
scripts/faceless.sh
Purpose-built end-to-end pipeline. Handles: create project â generate â poll â export â download URL.
faceless.sh --topic TEXT [--duration 30] [--voice rachel] [--style cinematic]
[--image-model ID] [--video-model ID] [--video-duration SEC]
[--storyboard-only] [--no-animated-video]
[--aspect 9:16] [--captions true] [--caption-style ID]
[--resolution 1080p] [--name TEXT] [--yes] [--dry-run]
Run bash scripts/faceless.sh --help for full usage.
scripts/blitzreels.sh
Generic API helper for ad-hoc calls. Requires BLITZREELS_ALLOW_EXPENSIVE=1 for expensive endpoints.
bash scripts/blitzreels.sh METHOD PATH [JSON_BODY]
Environment Variables
| Variable | Required | Description |
|---|---|---|
BLITZREELS_API_KEY |
Yes | API key for authentication |
BLITZREELS_API_BASE_URL |
No | Override base URL (default: https://www.blitzreels.com/api/v1) |
BLITZREELS_ALLOW_EXPENSIVE |
No | Set to 1 to allow expensive calls via blitzreels.sh |
API Endpoints
| Method | Path | Description |
|---|---|---|
| POST | /projects |
Create a new project |
| POST | /projects/{id}/faceless |
Start faceless generation (expensive) |
| GET | /jobs/{job_id} |
Poll job status |
| GET | /projects/{id}/context?mode=timeline |
Inspect generated timeline |
| POST | /projects/{id}/export |
Start export (expensive) |
| GET | /exports/{export_id} |
Get export status + download URL |
| GET | /voices |
List available voices |
| GET | /visual-styles |
List available visual styles |
| GET | /caption-styles |
List caption style presets |
Faceless Request Body
{
"topic": "5 morning routine tips",
"duration_seconds": 30,
"visual_style": "cinematic",
"voice_id": "21m00Tcm4TlvDq8ikWAM",
"captions": true,
"caption_style": "bold-center"
}
| Field | Type | Required | Description |
|---|---|---|---|
| topic | string | one of topic/script | AI generates script from this |
| script | string | one of topic/script | Exact narration text, paragraph breaks = scenes |
| duration_seconds | number | yes | 10â120 |
| visual_style | string | no | Style ID (default: cinematic) |
| voice_id | string | no | ElevenLabs voice ID (default: Rachel) |
| captions | boolean | no | Enable captions (default: true) |
| caption_style | string | no | Caption style preset |
| imageModelId | string | no | Image generation model (default: google-gemini-2.5-flash-image) |
| videoModel | string | no | Video generation model (default: kling-2.1) |
| videoDuration | number | no | Per-clip duration in seconds, 3â10 |
| storyboardOnly | boolean | no | Generate images only, skip video/audio (default: false) |
| generateAnimatedVideos | boolean | no | Set false to skip video animation (default: true) |
Generation Modes
| Mode | Flag | What’s Generated |
|---|---|---|
| Full video (default) | â | Images â video clips â voiceover â music â captions |
| Storyboard only | --storyboard-only |
Scene images only (no video, no audio). Cheapest option for previewing. |
| Images + audio | --no-animated-video |
Scene images + voiceover + music (slideshow-style, no video animation) |
References
references/models.mdâ All image and video generation models with costs and capabilitiesreferences/voices.mdâ All available voice IDs, categories, and descriptionsreferences/visual-styles.mdâ All available visual style IDs and descriptions
Safety
faceless.shprompts for confirmation before calling expensive endpoints (bypass with--yes)blitzreels.shrequiresBLITZREELS_ALLOW_EXPENSIVE=1for/facelessand/exportpaths- Always present cost/plan to user before generation
Notes
- Use
https://www.blitzreels.com/api/v1as your base URL.https://blitzreels.comredirects towww, and some HTTP clients drop theAuthorizationheader on redirects. POST /projects/{id}/facelessreturns aplanandinput_warningsto help you predict what will be generated.- Full OpenAPI spec:
https://www.blitzreels.com/api/openapi.json
Rate Limits
| Plan | Requests/min |
|---|---|
| Free | 10 |
| Creator | 30 |
| Pro | 60 |
| Agency | 120 |