Full AI agent template for creating, scheduling, and deploying visual content at scale
Category: Content Generation + Distribution Runtime: Claude Code AIOS Rating: 9/10
Complete end-to-end content creation pipeline. Takes brand files and product photos, generates images and videos using multiple AI providers, manages everything through Airtable, and schedules posts via Blotato across YouTube, TikTok, Instagram, and LinkedIn.
This is the most complete system in the collection — it extends R56's generation engine with campaign planning, social scheduling, brand voice files, and Modal deployment.
Brand Files → Agent reads context → Campaign Calendar (Airtable)
(BRAND.md) (AGENT.md brain) 30 days of content planned
Campaign → Generate prompts → Airtable review → Human approves
Approved → Provider Registry → Google / Kie / WaveSpeed → Airtable
routes to best (images + videos) (output)
available model
Content → Blotato scheduling → Multi-platform posting
(SKILL.md) with timezone handling
| Platform | Role | Cost |
|---|---|---|
| Google AI Studio | Primary image gen (Imagen 3) + video gen (Veo 3.1) | Free tier |
| Kie AI | Secondary provider — NanoBanana Pro, Kling 3.0 | Per-generation |
| Airtable | Content management — prompts, images, videos, calendar | Free tier |
| Blotato | Social media scheduling across all platforms | Paid |
| Modal | Serverless deployment for YouTube→LinkedIn automation | Free tier |
| Claude Code | Agent runtime — executes AGENT.md brain + skills | Subscription |
Plan and execute a full month of content from brand files.
- Brand Discovery → Content Calendar → Image Generation → Scheduling
On-demand image and video generation with cost controls.
- Read brand context → Write prompts (SEALCaM/BOPA) → Airtable review → Provider routing → Log results
Automated content repurposing pipeline deployed on Modal.
- Extract video metadata → Generate LinkedIn posts → Deploy as cron job
r57-template-community/
├── .agent/
│ ├── AGENT.md (404 lines — brain)
│ ├── skills/
│ │ ├── blotato_best_practices/SKILL.md
│ │ └── modal_deployment/SKILL.md
│ └── workflows/
│ ├── 30-day-campaign.md
│ ├── generate-content.md
│ └── youtube-to-linkedin.md
├── references/
│ ├── .env.example
│ ├── docs/
│ │ ├── prompt-best-practices.md
│ │ ├── fabric_BRAND.md
│ │ └── imma_BRAND.md
│ ├── inputs/ (product photos)
│ └── outputs/ (generated assets)
└── tools/
├── config.py
├── image_gen.py
├── video_gen.py
├── video_analyze.py
├── airtable.py
├── kie_upload.py
├── setup_airtable.py
├── modal_pipeline_example.py
└── providers/
├── google.py
├── kie.py
└── wavespeed.py
Required:
| Variable | Source | Purpose |
|---|---|---|
GOOGLE_API_KEY |
aistudio.google.com | Imagen 3 + Veo 3.1 |
KIE_API_KEY |
kie.ai dashboard | NanoBanana + Kling + Suno |
AIRTABLE_API_KEY |
airtable.com/create/tokens | Content management |
AIRTABLE_BASE_ID |
Airtable URL | Target base |
BLOTATO_API_KEY |
blotato.com → Settings → API | Social scheduling |
Optional:
| Variable | Purpose |
|---|---|
WAVESPEED_API_KEY |
Sora fallback provider |
| Generation | Model | Provider | Cost | Time |
|---|---|---|---|---|
| Image | Imagen 3 | Free tier | ~10-30s | |
| Image | NanoBanana Pro | Kie AI | $0.09 | ~30-60s |
| Video 5s | Veo 3.1 | Free tier | ~1-2 min | |
| Video 5s | Kling 3.0 | Kie AI | $0.28 | ~2-4 min |
| Video 5s | Sora 2 Pro | WaveSpeed | $0.50 | ~2-5 min |
Example: 30-day campaign (30 images + 10 videos) ≈ $5.50
- Show Costs Before Generating — Display estimate and get explicit confirmation
- Human Review at Each Stage — Prompts → Airtable → approval → generation → review
- Use Brand Voice Files — Always read *_BRAND.md before generating content
| Problem | Cause | Fix |
|---|---|---|
| Wrong provider for model | R57 defaults Kling to Kie AI | Check provider registry in tools/providers/init.py |
| Brand voice inconsistent | Missing BRAND.md file | Create comprehensive brand file before generating |
| Blotato timezone issues | Server vs user timezone mismatch | Use Blotato's useNextFreeSlot for auto-scheduling |
Built by Copyweb — architecting how marketing teams operate in an AI-first world.