NEEV · 🏛️ Command Center

Content Engine

Open Content Review → live pipeline, review gates, no Baserow needed

The finalized content pipeline — one topic fans into every format, manual and automation versions compete, the winner feeds back into what gets planned next. Same diagram as the Content Creation deep-view, promoted here as its own page for direct access.

live / proven
gap — not built yet
own pipeline — bridge needed
Production Pipeline — full diagram
Every step of "Production → Publish" below, broken out: the real tools (ComfyUI+FLUX, Remotion, Gemini TTS), real file paths, and the 6 Baserow tables in between.
→
Imran Shabiha Gulfisha Ada 4 profiles — any of them can create owns Channel 01_channels · 936 — live feeds an idea into Topic Master — NEW TABLE pillar · cluster · planned_formats 1 topic → many pieces, tracked as one family fans into formats bridge field needed Image post single · auto-publish Carousel 8-slide + LinkedIn PDF Reel 9:16 video + voiceover 939 · content_master, all 3 Blog post 761 every format, same rule ↓ Manually created Shabiha · ChatGPT/Grok/CapCut Auto-generated Flux + Remotion pipeline who MADE it — needs field: created_mode (939). Not the same as publish mode below ↓ Asset Library (reuse) 942/943 — ChatGPT = images only pulls reusable asset Production → Publish 941–948 · Flux + Remotion + TTS → Meta proven live full diagram, tools & files → Interactions / Leads 950 / 951 Performance 949 — proven live compares — which version wins Live page arisegulf.com / islamiczindagi.com own pipeline — not touched

Source of truth: same diagram as Command Center → Arise Enterprise → Shabiha/Imran/Gulfisha/Ada → Content Creation. Kept in sync by hand — update both places if this changes.

Publish policy — locked, per format × platform

Image post (single)
IG + FB: auto
Verified live 8 Aug — graph.instagram.com / graph.facebook.com direct.
Carousel (8-slide)
IG: manual only
Graph API silently downgrades an IG carousel to a single image — tested 4x, confirmed broken. Posted by hand from the app, every time.
Carousel (8-slide)
FB: auto
attached_media works — Facebook carousel publishes automatically.
Carousel → LinkedIn PDF
manual only
Composed (856KB PDF) but uploaded by hand — no LinkedIn auto-publish path built.
Reel (video)
IG + FB: auto
LinkedIn stays manual for video too — no automated path built there yet.

This is a different axis from "Manually created / Auto-generated" in the diagram above — that's who made the piece (Shabiha by hand vs. the Flux+Remotion pipeline). This table is how it gets posted once it's ready, and it's the one that's actually tested and locked — not a gap.

Channels this pipeline serves

Brand channels
Arise Gulf Growth · Islamic Zindagi · Nek Zindagi · Pure Hearts · Al-Samd · Alina (8yo persona channel) — all logged against 01_channels, table 936.
Owners
Any of the 4 profiles (Imran, Shabiha, Gulfisha, Ada) can own a channel and create for it — Shabiha is the confirmed brand face + full-pipeline owner for Islamic Zindagi Arise today.

Known gaps — noted, deferred by design

939 multiple rows per idea
bug
One content idea currently creates several rows in content_master instead of one row holding script + image prompts + everything together. An earlier planned version didn't have this — flagged as a real regression to investigate, fix deferred on purpose (not blocking today's page fix).
Topic Master — new table
gap
pillar / cluster / planned_formats — the table that lets one topic fan into carousel/reel/image as one tracked family. Not built yet.

Build status — detail

Channel01_channels · table 936
live
Any of the 4 profiles can own a channel — proven, live.
Topic Masternew table — not built
gap
The piece that lets one topic fan into image/carousel/reel as one tracked family (pillar/cluster/planned_formats) doesn't exist yet — content pieces aren't linked back to a shared topic today.
Image post / Carousel / Reeltable 939, content_master
live
All 3 real formats live and tested (image post + carousel + reel — see Publish Policy above for each one's real per-platform rule). "Manually created / Auto-generated" is a separate, still-open axis below.
Manually created vs Auto-generatedneeds created_mode field, row-level, 939
gap
Both paths exist in practice (Shabiha manual via ChatGPT/Grok/CapCut; automation via the Flux+Remotion pipeline) but there's no field yet recording which mode produced a given row — needed for the Performance stage to actually compare "which version wins." Not to be confused with Publish Policy above (locked, already tested).
Asset Library (reuse)942/943, + 940 content_variants
live
Reusable image/B-roll/asset library — live. ChatGPT covers images only, not the full asset range yet.
Production → Publish941–948 — full diagram ↗
live
Prompt Library → Generated Assets → Voice Jobs → Remotion Components → Render Jobs → Publish (947 → 948), unchanged, proven. One card on the diagram above — see Publish Policy above for the real per-platform rules.
Performance949
live
Proven live — this is the stage that's meant to feed the manual-vs-automation comparison back upstream, blocked on the created_mode field above.
Interactions / Leads950 / 951
live
Live.
Blog post761
bridge needed
Exists as its own pipeline, but has no field linking it back to Topic Master — a topic that becomes a blog post isn't tracked as part of the same content family yet.
Live pagearisegulf.com / islamiczindagi.com
own pipeline
Deliberately not touched by this pipeline — separate system.