# Brief 8 — Synthesis: cross-post tool + video for an agent-run BIP account
**Role:** cross-cut of tools (a) + video (b) + payoff (c)  
**Context:** one founder, one agent (Claude/APIs/MCP), usectl self-host easy, X = home, text + occasional cover image, no video yet  
**Written:** 2026-09-05 (Asia/Tbilisi)  
**Inputs:** `/workspace/brief-8-tools.md` · `/workspace/brief-8-video.md` · `/workspace/brief-8-payoff.md`  
**Deliverables:** `brief-8-tools`, `brief-8-video`, `brief-8-payoff`, `brief-8-synthesis`
---
## 1. Executive summary
1. **Tool:** For an agent that must eventually hit X + YT + TikTok + LI + Threads + Bluesky + IG + Reddit + Mastodon, **Postiz self-hosted on usectl** is the only researched publisher that covers **all nine** with REST + MCP + CLI. Cloud fallback: **Zernio** (ex-Late; API/MCP; X pass-through at official rates; no Mastodon) or **Buffer Essentials** (MCP; X bundled; no Reddit). **Hypefury lost X** — skip. Typefully is excellent for **X-writing + MCP** but is **not** a nine-network publisher (no YT/TikTok/IG/Reddit).  
   Sources: tools §exec + tables A/B · https://docs.postiz.com/public-api/introduction · https://zernio.com/pricing · https://buffer.com/api · https://hypefury.com/features-pricing/
2. **X cost:** Official pay-per-use is **$0.015** per Post Create and **$0.200** with a URL — BIP ship posts are almost all linked. Who pays: Mixpost/Ayrshare/Zernio = **you** (BYO or pass-through); Buffer/Typefully/Publer paid ≈ bundled; Metricool = **+$10/mo per X**; Postiz cloud absorb = **unconfirmed**.  
   Source: https://docs.x.com/x-api/getting-started/pricing.md · tools §X compatibility
3. **Video:** Successful short-form BIP in 2026 is problem hooks, short face, screen+VO, slideshows — **not** polished studio. Hard-number small-base wins (Natural Write, Flowfy) are **consumer / TikTok-native apps**, not infra SaaS. Text BIP still compounds without video (Levels ~1B X views; Marc Lou text MRR). Faceless stock+TTS = **YouTube YPP / inauthentic risk**. **Verdict: after X baseline, not a daily TikTok grind.** Thin test = ~2h/week Loom→Shorts.  
   Sources: video §1–5 · https://levels.io/1-billion-views-on-x-in-12-months · https://blog.youtube/news-and-events/improving-ai-labels-viewers-creators/ · https://support.google.com/youtube/answer/1311392
4. **Cross-post payoff:** Public UTM’d same-post → 3-network customer splits are rare. Pattern that repeats: **adapted** presence on 3–4 text networks helps lists; **customers lean LinkedIn + Reddit (+ X for PLG speed)**; PH/HN/short-form win **views** more than paying users. Identical caption blast hurts.  
   Sources: payoff §1–3 · https://www.teract.ai/resources/linkedin-vs-twitter-b2b-2026 · https://justneeda.com/articles/video-marketing-2026 · https://monolit.sh/blog/social-media-cross-posting-strategy-what-to-post-where-2026
**Bottom line:** Ship **Postiz on usectl** (or Zernio/Buffer if cloud-only). Start networks **X + LinkedIn + Bluesky (or Threads)**; Reddit as participate-don’t-blast. **No YT/TikTok until ~90 days X (and ideally LI) baseline with UTMs.** One story → native adaptations, staggered; never identical dump.
---
## 2. Reconciled tool ranking (tools matrix wins)
| Rank | Tool | Why (merged) | When to pick instead |
|------|------|--------------|----------------------|
| **1** | **Postiz (self-host on usectl)** | Only full 9-network matrix + API/MCP/CLI; matches “self-host easy” | Cloud ops aversion → Zernio or Buffer |
| **2** | **Zernio (Late)** | Best pure API/MCP DX; 2 accounts free; X at official pass-through | Need Mastodon in one box → Postiz or Buffer |
| **3** | **Buffer Essentials** | Lowest-friction cloud MCP; X included; escape hatch to YT/TikTok later | Need Reddit in-box → Postiz or Zernio |
**Demoted vs payoff draft:** Typefully stays **X-writing home / MCP drafting companion**, not the cross-post publisher of record (platform gaps). Payoff’s draft Typefully→Postiz→Buffer is superseded by tools matrix for the nine-network + agent requirement.
**Skip:** Hypefury (no X), SocialBee (weak API), Ayrshare ($149 floor for one brand).
---
## 3. Network plan + cadence
**Phase 1 (now):** X home · LinkedIn customer channel · Bluesky or Threads as cheap third · Reddit comments/“what I learned” only  
**Phase 2 (~90 days):** Reassess YT Shorts (one weekly Loom→clips) if X/LI UTMs show demand  
**Defer:** daily TikTok, multi-account farms, Arcads/HeyGen volume, Fastlane-style 100 posts/week
**Weekly (agent-adapted, one story):**
- X: 5–10 posts/wk (incl. threads + reply-heavy day)
- LinkedIn: 3–5/wk (longer frame; stagger ~24h after X or LI-first then X)
- Bluesky/Threads: 2–4/wk short native cuts
- Reddit: quality comments, not identical blasts
- UTM every link (`utm_source=x|linkedin|bluesky`)
Sources: payoff §4 · Conbersa multi-platform · Monolit cadence · JustNeeda under-$5k X-first
---
## 4. Video decision (merged)
| Option | Hours | Fit for usectl | Call |
|--------|------:|----------------|------|
| Daily TikTok / faceless AI farm | 8–15h/wk or spam risk | Poor (infra buyers ≠ Flowfy/Natural Write) | **No** |
| Weekly Loom/screen+VO → 2–4 Shorts | ~2h/wk | Good thin wedge after X baseline | **Later / optional test** |
| Stay text+image on X/LI | current | Matches Levels/Marc Lou path | **Default now** |
AI as **editor/clipper on your footage** = OK. AI as **interchangeable stock+TTS factory** = YT demonetization / trust risk for developer buyers.
---
## 5. Audience question (one line)
when you last tried a new infra or self-host tool, what actually made you click signup — an x thread, a linkedin post, a reddit comment, or something else?
---
## 6. Recommended next actions (weebo / Claude)
1. Deploy **Postiz** on a usectl machine; connect **X + LinkedIn + Bluesky/Threads** first; wire agent via MCP/API.  
2. Instrument **UTMs + “how did you hear?”** for 90 days.  
3. Keep **Typefully** optional for X draft craft if writing UX matters; do not rely on it for YT/TikTok/Reddit.  
4. Budget X link posts at **$0.20** each if BYO/pass-through (Zernio/Mixpost/Ayrshare/self-host).  
5. Video only after phase-1 baseline; then OpusClip/Descript on human Loom, Shorts-first.
---
## COUNTS (synthesis layer)
| Metric | Count |
|--------|------:|
| Angle docs merged | **3** |
| Top tools after reconcile | **3** (Postiz, Zernio, Buffer) |
| Phase-1 networks | **3 + Reddit participate** |
| Hard “no” calls | **2** (Hypefury; daily TikTok now) |
| Confirmed open gaps carried forward | **Typefully $ scrape · Postiz cloud X absorb · no UTM’d 3-network customer diary · consumer-bias video cases** |
---
## Missing / still open
- Postiz cloud: does subscription absorb X API? (tools gap)  
- Typefully exact list prices (JS pricing page)  
- Independent same-story → X/LI/Bluesky customer attribution for infra SaaS  
- usectl buyer channel survey  
- Live re-check Captions/Arcads/Revid pricing before any video tool spend  
---
*Research only. Inbox docs: brief-8-tools, brief-8-video, brief-8-payoff, brief-8-synthesis. Threadbus outcome posted separately.*