Owner operations
The whole desk, in one place.
App health, production, publishing and results. Agents maintain the records; this is your view of the work.
Issues to work through
Measurement coverage
Native and direct measurements stay separate from Buffer reports. Buffer reports can refresh daily; sent status does not verify native publication. Stories and feed posts keep separate counters.
Coming up
Confirmed schedules and work in progress.
Latest results
Native observations with a capture time.
Activity
Evidence of what happened across the desk.
Publishing calendar
All platforms, with scheduled and actual publication times in Eastern time.
Content performance
Every video, Reel, post and article. Open a record for its history and source details.
Views, unique viewers and reach are different measures. Rates use the displayed denominator. Missing data is shown as unavailable; counts from different platforms are kept separate. History begins at the first recorded observation.
Growth
Fast reads and slow trends, each labelled by how far it can be trusted yet. Computed from the collected measurements; nothing here is entered by hand.
This week against last week
Rolling seven-day weeks ending at the latest capture. Account counters and per-post counters stay on their own platform definitions.
Signals
Plain flags from fixed rules, labelled by speed. They point at something to look at; they are not advice.
Matched-age race
Each piece against comparable pieces on the same platform and format at the same age. Long videos say nothing about reach before 72 h.
Long-video diagnosis
Same-video reads: the opening, the biggest drops with what was on screen and said, traffic and conversion.
Shorts & Reels
YouTube and Instagram columns never combine. Rates below the minimum sample show their raw counts. Shorts views count every replay, so subscribers per 1,000 use YouTube’s engaged views. A dash is not a zero: hover it for the reason (Analytics rows lag one to two days; stayed-to-watch is Studio-only).
TikTok & Facebook
Numbers read from TikTok Studio and Meta Business Suite at 24 h, 72 h and 7 days. Each platform keeps its own definition; nothing here combines with YouTube or Instagram. Posts not yet verified on the platform are aged from Buffer's exact sent time and say so.
Strengths and weaknesses by tag
For each tag value, the median of its pieces against the rest of the same class at one matched age. Fewer than 3 pieces on either side is a hint only. Tags overlap (most episode excerpts are also cloned voice and cartoon), so a rule here is a pattern to test, not a cause.
Experiments
Growth Lab tests with their win criterion and matched-age reads. The Growth Lab log keeps each decision.
App usage
Product activity and waitlist growth, alongside your content results.
Recorded activity · last 30 days
UTC days. Accounts with recorded actions measure writes, not sessions or all active users. Empty days are not reconstructed.
Staging contains test activity. Production is reported separately when provisioned. Research cost already includes search. Metered samples and report latency retain their coverage counts. Customer identities, emails and collection contents stay out of this workspace.
Production & agents
Current work, existing workflow owners, dependencies and the next concrete step.
Open collection issues
Sessions revisit these records until recovery is verified. Other sources keep collecting.
Production work
Agent interface
A versioned API, immutable measurements and revision checks give Codex and Claude the same operating record.
node ops/cli.mjs contract node ops/cli.mjs state node ops/cli.mjs apply /absolute/path/to/bundle.json node ops/cli.mjs history ig:DenPRqhPDhJ
An operations record does not grant permission to publish, spend, change an account or access another scope. Each workflow retains its existing approval and access rules.
Brand press & media access
Official contacts, requests and confirmed access. Agents keep this register current.
A sent request is separate from press-list membership and granted media rights. Public files still carry their own credits, usage rules and embargoes.
Media & shooting
Set up the camera for your next shot, or plan and check a watch's coverage.
1. Where is your camera now?
Choose the setup you have applied. If you're unsure, choose “Not sure” for the full setup.
This page uses what you select; it cannot read or change your phone.
2. What do you want to shoot next?
2. Choose this watch shoot
Preview only. Codex or Claude can register this individual watch and its sample terms when you're ready to save footage.
Watch features · leave anything uncertain as unknown
3. Work through one setup at a time
You can jump to any batch. Settings always start from the current setup above. Use “I've applied this setup” after changing your phone.
Coverage and source files
Codex and Claude update coverage after reviewing the exact files.
Content reuse review dates
Review published Instagram Reels at 30, 60 and 90 days. Compare performance and consider a fresh edit before approving another post.
These are review dates. Nothing is scheduled to repost automatically.
App & data health
Collection health, collector coverage, freshness and missing permissions.