Content Engine — Final

Multi-business content engine for the Autono Labs portfolio. Anti-AI-feel by default, hooks-first for scroll-stopping, voice-tuned per brand, content-mix-tracked, AI-tell monitor keeps it current.

Anyone can clone this repo, follow ONBOARDING.md, and have a brand publishing within 2 weeks.

🔗 Repo: https://github.com/AutonoLabs/content-engine-final 📖 Start here: ONBOARDING.md 🐛 Issues: https://github.com/AutonoLabs/content-engine-final/issues

---

What this is

A complete content engine + publishing pipeline for any brand:

1. Voice layer — generates a brand-specific voice profile from a brief (no seed posts needed)
2. Hook + narrative + visual layer — platform-specific patterns that stop scroll
3. Anti-AI layer — banned phrases + quarterly AI-tell monitor
4. Claim-verification layer — sector-aware (FCA, FDA, FTC, state bar)
5. Diversity layer — rotation enforcement across format / theme / style / hook / visual / audience
6. Publishing layer — Blotato integration, multi-platform scheduling
7. Tracking layer — content-mix + performance-log in markdown
8. Skill layer — four generalizable skills that work for any new business

---

Quick start (60 seconds)

git clone https://github.com/AutonoLabs/content-engine-final.git
cd content-engine-final
pip install -r requirements.txt
cp .env.example .env  # Fill in keys (see docs/ENV-WIRING.md)

# Onboard your first brand
python scripts/onboard_brand.py --name mybrand --sector fintech

# Fill in brand-brief.md, then run voice-from-brief skill
# Read ONBOARDING.md for full flow

---

What's in the box

content-engine-final/
├── ONBOARDING.md                    # Clone → publish in 2 weeks
├── README.md                        # This file
├── LICENSE                          # MIT
├── .env.example                     # API keys template
├── .gitignore
├── requirements.txt
├── docs/                            # Architecture + rules
│   ├── platform-hooks.md
│   ├── visual-hooks.md
│   ├── audience-demographics.md
│   ├── narrative-styles.md
│   ├── anti-ai-feel.md
│   ├── ai-tell-history.md
│   ├── content-type-taxonomy.md
│   ├── diversity-rules.md
│   ├── content-pull.md
│   ├── publish-runbook.md
│   ├── media-brief.template.md
│   ├── voice-profile.template.md
│   ├── ENV-WIRING.md
│   ├── WEEKLY-BATCH-FLOW.md
│   └── TROUBLESHOOTING.md
├── prompts/                         # Per-platform generation prompts
│   ├── linkedin.md
│   ├── x.md
│   ├── instagram.md
│   ├── tiktok.md
│   ├── youtube-shorts.md
│   └── threads.md
├── skills/                          # Generalizable skill layer (works for any new business)
│   ├── voice-from-brief.md
│   ├── ai-tell-monitor.md
│   ├── claim-verifier.md
│   └── brand-adapter.md
├── brands/                          # Per-brand folder templates
│   ├── voice-profile.template.md
│   ├── target-audience.template.md
│   ├── verified-facts.template.md
│   ├── content-mix.template.md
│   ├── performance-log.template.md
│   ├── brand-brief.template.md
│   └── blotato-accounts.template.md
├── scripts/                         # Operational pipeline code
│   ├── blotato_client.py            # Blotato API wrapper
│   ├── blotato_list_accounts.py
│   ├── blotato_publish.py
│   ├── blotato_get_post_status.py
│   ├── higgsfield_generate.py       # Video generation (6 models)
│   ├── content_pull.py              # Multi-source extraction (playwright → blotato → curl)
│   ├── performance_pull.py          # Engagement tracking
│   ├── validate_post.py             # Anti-AI + diversity + claim checks
│   └── onboard_brand.py             # New-brand bootstrap
└── examples/                        # Reference brand
    └── allsquared/
        ├── README.md
        ├── voice-profile.md
        ├── target-audience.md
        ├── verified-facts.md
        ├── content-mix.md
        └── weeks/2026-W28/
            ├── captions.md
            └── media-briefs.md

---

What does it actually do?

For each brand in the portfolio:

1. Generate voice profile from brand brief (no seed posts needed) — skills/voice-from-brief.md
2. Plan weekly batch with diversity enforcement (rotation across format / theme / style / hook / visual / audience / platform) — docs/diversity-rules.md + docs/WEEKLY-BATCH-FLOW.md
3. Draft captions using platform-specific hooks + narrative styles — prompts/<platform>.md + docs/platform-hooks.md + docs/narrative-styles.md
4. Verify claims against brands/<brand>/verified-facts.md (sector-aware: FCA / FDA / FTC / state bar) — skills/claim-verifier.md
5. Validate post against anti-AI + diversity rules — scripts/validate_post.py
6. Generate or drop media via Higgsfield or manual-media mode — scripts/higgsfield_generate.py
7. Publish via Blotato across 7+ platforms — scripts/blotato_publish.py
8. Track performance + diversity — scripts/performance_pull.py + brands/<brand>/content-mix.md + performance-log.md
9. Refine voice over time via edit pass — brands/<brand>/exemplars/edits/
10. Quarterly AI-tell refresh — skills/ai-tell-monitor.md

---

Why it's different

vs. plain LLM use:

• Anti-AI banned phrases enforced + AI-tell monitor keeps it current
• Voice profile converges to actual brand via edit pass
• Claim verification prevents stat hallucination
• Diversity rules prevent boring sameness

vs. other content tools:

• Multi-business, multi-platform, multi-format (16 content types)
• Works for any new business without seed posts
• Sector-aware compliance (health, finance, legal, professional)
• No vendor lock-in (markdown + scripts, not SaaS)
• MIT licensed, fully public, fork-friendly

---

Architecture principles

1. Markdown-first. All knowledge in plain .md files. Version-controlled, fork-friendly, no proprietary format.
2. Voice from brief, not from posts. New brands don't need seed posts to start.
3. Anti-AI by default. Banned phrases + structural patterns. Quarterly refresh.
4. Verified facts are HARD. No unverified claims in posts.
5. Diversity enforced. No boring sameness.
6. Edit pass is the loop. Voice converges over time.
7. Multi-format. 16 content types, rotation enforced.
8. Multi-platform. 7+ platforms supported via Blotato.
9. Scripts over SaaS. Run your own pipeline.
10. Fork-friendly. MIT license, public repo, anyone can use.

---

Tested & working

✓ python scripts/validate_post.py — anti-AI catches banned phrases
✓ python scripts/onboard_brand.py --help — onboarding ready
✓ python scripts/higgsfield_generate.py --list-models — 6 models available
✓ python scripts/blotato_list_accounts.py — account discovery
✓ All 9 scripts parse cleanly

---

Sector support

Sector	Sector rules in claim-verifier	Brand examples
Consumer health (FDA/FTC)	✓	Yapper
Fintech (FCA/SEC)	✓	AllSquared
Legal (state bar)	✓	(future)
Professional services	✓	(future)
Education / creative	✓	TreeAI, etc.
Other	✓	any

---

Cost (per brand, monthly)

• Blotato: $19-$99 (depending on plan + account count)
• Higgsfield Ultra: $99/mo (6000 credits, all models)
• Optional: OpenAI/Anthropic/Perplexity for research ($10-$30)
• Total: ~$120-$230/mo per brand

Manual-media mode (skip Higgsfield) drops it to ~$20-$100/mo.

---

Roadmap

• Cron job scheduler for weekly themes (in repo)
• Discord approval integration (messaging-based edits)
• Per-platform engagement analytics dashboard
• A/B testing framework (rotate 2 versions of same theme)
• Auto-translation for multi-language brands

---

Support

• New to this? Start with ONBOARDING.md
• Adding a brand? Read skills/brand-adapter.md
• Stuck? Check docs/TROUBLESHOOTING.md
• Want to extend? Read docs/ and skills/, contribute back via PR

---

License

MIT — see LICENSE. Fork-friendly, use it, modify it, ship it.

---

Built by: Autono Labs, 2026-07-08 Maintainer: Autono Labs Status: Active, v1.0