Publishing is the easy part. Social tracks what happened after: engagement rate and half-life per post, the 23 publishes that failed and why, and the 61 inbox items aging without an owner.
A clip that dies in 6.8 hours and one still earning at 14.2 are the same format on paper. Half-life is how you tell them apart.
social.cedxsystems.com — live build
Runs on demo data. The accounts, posts and metrics are the demo workspace's.
320 posts in the library129 published · 54 scheduled
125K engagements trackedaverage rate 5.2% on published
9.9h median half-lifedecay measured per post, not per hunch
What it is
Three things social operations has to get right.
The calendar is the queue
Month grid, week list, day detail — every cell a real post with its network and format marked. 54 scheduled posts sit ahead; draft, published and failed are filters of the same view, so the queue and the record are never two places.
54 scheduled upcoming
Filter by network and by status
Compose straight from a cell
social — screen-2
Posts with impressions, not vibes
321 rows with status, network, format, impressions, engagements, rate and owner. The top carousel earned 4.5K engagements on 51K impressions — an 8.8 rate. The best clip holds a rate of 10 on fewer impressions. Both truths are on the same table.
Impressions, engagements and rate per post
Owner on every row
CSV export of the filtered list
social — screen-3
One brief, nine variants, bands attached
The composer takes a brief and writes three caption lanes — bold, helpful, data-led — per network, nine variants for three networks. Each carries the predicted engagement band for its format and network before anyone commits to it.
3 lanes × 3 networks = 9 variants
Predicted bands: stories 4.2–7.1%, feed 2.4–4.1%
Repurpose top posts one-to-many
social — screen-4
Product tour
Four screens, captured from the running build.
Not a mockup and not a concept deck. This is what opens at /app/social.
social.cedxsystems.com
01 — Overview
Failures are first-class news
The alerts lead with what broke: 23 failed publishes — scheduled video stuck in the failed state; seven posts dying faster than their format's median; one clip at −78% against its expected curve, and another at +71% worth reposting.
Root-caused alerts, publish and decay
Social health 68% computed on screen
Engagement total 125K from 129 published
02 — Calendar
See the month like a newsroom
Launch cuts, how-to cards, hiring proofs and recap frames share one grid — with the failed cell as visible as the published one, because that is where the work is.
Month grid, week list, day detail
Network and format marked per cell
Focus day with one click
03 — Posts
The library keeps the receipts
Draft 55, in review 34, scheduled 54, published 129, failed 23 — the status chips are counts, not labels. Open any row and impressions, engagements, rate and owner travel with it.
321 posts matching the view
Rate up to 10 on the best clip
Status chips double as filters
04 — Composer
Bands before you bet
The predicted-bands rail prices format by network — video 3.6–6.4%, community 3.1–5.2%, professional 1.9–3.4% — and each variant quotes its band. Send to drafts lands the pick in the post library.
Best predicted band 7.3% on this brief
Five repurpose sources flagged eligible
Zero sent to drafts until a human picks
Who runs it
Three roles keep the channel honest.
Roles, not references. We have no named customers yet, so nobody in these photographs is quoted, credited or claimed as one.
Social manager
Owns the calendar and the composer, and picks the variant whose band justifies the post.
320 posts · 54 scheduled
Content producer
Cuts the clips and watches half-life — a 6.8-hour clip gets reworked, a 14.2-hour one gets reposted.
median HL · 9.9h
Performance lead
Reads decay next to paid: when organic dies faster than baseline, the boost decision has a number behind it.
decay alerts · 85 anomalies
The shape of it
What the demo workspace actually looks like.
Every figure below is legible in the captures above. Nothing here is a projection of your estate — it is the state of the demo data.
129posts publishedof 320 in the library
5.2%average engagement rate125K engagements in view
61inbox items unassignedmentions aging without an owner
23failed publishes in queueroot-caused on the pulse
Posts by status320 posts across the queue
Published129
Draft55
Scheduled54
In review34
Failed23
Predicted engagement bands by networkupper bound · format × network baselines
Stories4.2–7.1%
Video3.6–6.4%
Community3.1–5.2%
Photos2.8–4.6%
Feed2.4–4.1%
Professional1.9–3.4%
Social healthpublish + decay, inputs on screen
68%
Published 129 · eng rate 5.2%
Failed 23 · half-life 9.9h
How it runs
A post's life, in the order it actually happens.
01
Plan
The calendar holds the month — every cell a real post with network and format, nothing floating in a separate doc.
02
Compose
One brief becomes nine variants across lanes and networks, each quoting its predicted band.
03
Publish
Scheduled posts go out — and the 23 that fail surface with causes instead of vanishing.
04
Decay
Half-life is measured per post; anomalies at −78% and +71% both become decisions, not mysteries.
One record
Social reads the record the estate already keeps.
Posts are not orphaned content. The assets, campaigns and analytics behind them are the estate's own.
Finding this out on the third call is worse for you than reading it here, and worse for us.
Social is not generally available. What opens today is the live build running on demo data.
The networks shown are generic — feed, photos, video, professional, community, stories. Which specific social platforms a given plan connects is not something this page claims.
Predicted bands are baseline estimates for format and network, labelled as such on screen. They are not performance promises.
We have no named customers to show you, so this page shows none. The account and posts in the captures are the demo workspace's.
No audit or compliance certification has been issued for Social. What we can evidence about hosting, encryption and access is on the security page.
Yes. Every screenshot is a capture of the running build and you can open the same build at /app/social. It runs on demo data — the accounts and posts are the demo workspace’s.
What is half-life?
The time a post keeps earning engagement. The demo median is 9.9 hours; posts dying faster than their format's median get flagged, and outliers in both directions become decisions — rework or repost.
What happens when a publish fails?
It stays visible. 23 failed publishes sit in the queue with a cause — scheduled video stuck in the failed state — instead of disappearing from the record.
Does the composer pick the variant for you?
No. It writes nine variants with predicted bands and sends nothing to drafts until a person picks — the sent counter reads zero on screen.
Is Social audited or certified?
No certification has been issued. What we can evidence about hosting, encryption, tenant isolation and production access is written up on the security page.
The queue is running. Go and look at it.
Live build, demo data, no card. Then ask which of your posts would show a 6.8-hour half-life.