Back Office · office.temerarii.xyz
Owns what the work looks like and what it says.
Seat · competed for

Owns what the work looks like and what it says.

Reports to the Executive Chairman. Titled Creative Lead; joined against the rest of the system on the track name Creative Lead.

What this seat owns

Pipelines it owns

Joined from governance/pipelines.json where owner_role is Creative Lead — 1 of 13. Ownership here means responsible for it existing and working, which is not the same as deciding what it produces.

PipelineBrandStatusPriorityOn air in (block lead)
Storyboards, scripts and style directionEvery brandpartial1BRIEF (Executive Manager), CROSS (counterparts), DEMO (each participant)

Pipelines it supports

Named in supporting_roles on 6 pipelines. Supporting is not owning. The owner is accountable for the thing working; this seat is accountable for its own contribution to it, and the distinction is what stops a stalled pipeline having two people who each thought it was the other's.

PipelineBrandStatusPriorityOn air in (block lead)
Universal brand baselineEvery brandpartial2BUILD (Technical Operator), CROSS (counterparts)
Curation and repost engineEvery brandmissing7PUBLISH (Technical Operator)
PDP tile GIF engineLil Shop of Culturemissing9BUILD (Technical Operator)
Virtual lookbook engineLil Shop of Culturemissing10BUILD (Technical Operator), DEMO (each participant)
Bully mascot worldsLil Shop of Culturemissing11BUILD (Technical Operator)
Any-brand intakeEvery brandmissing12BRIEF (Executive Manager), ADJUST (Executive Manager)

Episodes it leads

Zero. No episode in the season spine names Creative Lead as its led_by. The count is genuinely nil and is printed as nil.

Read that carefully, because zero here does not mean absent from the show. The spine names one lead per weekday: BRIEF and ADJUST to the Executive Manager, BUILD, GATE and PUBLISH to the Technical Operator, CHECKPOINT to the Producer. The remaining two days are attributed to everyone rather than to a seat — Wednesday CROSS reads counterparts and Friday DEMO reads each participant. Wednesday is this seat's day on air, and it is deliberately shared: the crossover exists so a creative runs a build and a technical writes a brief, both badly, in public. So the honest statement is that the Creative Lead is on air every single week and leads zero episodes by name — and only reading the spine tells you that.

The brands, and what each needs from this seat

The company runs 4 brands, and every one of them is listed — including the ones this seat has no pipeline on, because “nothing here” is an answer and an omission is not. This seat is on the hook for 2 of the rows below. Every brand is not a brand: it means the pipeline is brand-agnostic and serves all of them at once, so a failure there is felt everywhere simultaneously.

BrandOwnsSupportsThis seat / all pipelines
Finessionalsnonenone0 of 3
Temerarii Medianonenone0 of 1
Dom Davis Networknonenone0 of 1
Lil Shop of CulturenonePDP tile GIF engine, Virtual lookbook engine, Bully mascot worlds3 of 4
Every brandStoryboards, scripts and style directionUniversal brand baseline, Curation and repost engine, Any-brand intake4 of 4
The content goals per brand are not repeated here. Follower and revenue targets, the never-delegate list and the per-brand self/tech/creative split live on brand operations, which owns them. A second copy of a target is a second target — and the first time one is edited, two pages start telling a participant different things about what they are driving at.

What each one teaches, and how it is assessed

Every pipeline carries what it is meant to teach and the method it is judged by. This is the assessment that is not a point value — the seat is a paid training position, and this is the syllabus.

PipelineTeachesAssessed by
Storyboards, scripts and style direction
  • storyboarding as a deterministic input, not an inspiration
  • the technique menu is composable - one shape of content forever is the failure mode
  • AI-seeded is fine; unadjusted is not
verify_vo_caption_match plus the brand rails gates

What this seat is NOT responsible for

This is the section that prevents the arguments. Almost every dispute about a role is a dispute about its edge, not its centre — two people both believing something was the other's, or one person quietly annexing a decision that was never theirs. Writing the edge down is cheaper than settling it later, on air.

Requirements

Conditions of the seat. They are stated up front because discovering one of them in week three is a wasted seat for everybody.

Disciplines

Standing rules for how this seat works — not skills, but positions already taken, so they are not re-argued every sprint.

What you are measured on

20 of the 23 scoring rules apply to this seat — 19 that apply to everyone and 1 written for this track specifically. They split by who observes them: 14 are read by the machine and 6 are judged by a human. That split is the point of the format — the record is read from the system, not from anyone's recollection of your week.

ONE POINT = ONE US DOLLAR against the weekly adjustment schedule. The adjustment schedule these settle against is on the operations policy, and the weekly cap below limits what any single week can cost.

Automatic — the machine scores these

Nobody types these in. An emitter already in the stack produces the record, which is why each one names the file that proves it. 9 of 14 have a live emitter today.

WhatEffectRead from
The full gate suite green at the end of a week
every seat
not setengine/goal.py
prints GOAL ALIGNED
A blocker open more than three working days without its owner acting
every seat
not setengine/gates/verify_projects.py
prints PROJECTS PASS
A deliverable passed its target week without acceptance (owner seat)
every seat
not setengine/gates/verify_projects.py
prints PROJECTS PASS
A filed report not reviewed by the next standup (reviewer seat)
every seat
not setengine/gates/verify_eod_reports.py
prints EOD-REVIEW PASS
A missed calendar deadline
every seat
-30engine/gates/verify_calendar_alignment.py
prints CALENDAR-ALIGN PASS
A missed or postponed KPI deadline agreed in sprint planning
every seat
-20no emitter yet
An active deliverable with no acceptance criteria
every seat
not setengine/gates/verify_projects.py
prints PROJECTS PASS
Exposing an API key, token or secret in the repository
every seat
-100engine/gates/verify_no_secrets.py
prints NO-SECRETS PASS
Late to an obligatory meeting, live broadcast or event
every seat
-50no emitter yet
Missing end-of-day report — completed, blocking, next
every seat
-10engine/gates/verify_eod_reports.py
prints EOD-REPORTS PASS
No acknowledgement of a direct communication within 4 hours of an operational window
every seat
-15no emitter yet
Pushing unapproved or un-QA'd work to a live production surface
every seat
-50no emitter yet
Submitting work that ignores the documented baseline
every seat
-25engine/gates/verify_render_ready.py
prints RENDER-READY PASS
Work unreachable because it was never pushed, while the owner is offline
every seat
-30no emitter yet
5 of these cannot be scored today. They are real rules with no emitter yet — nothing in the stack currently writes the record they would be read from. They are listed because an admitted gap is safer than an invented signal: a rule pointed at a file nobody writes scores nobody, silently. Expect them to go live during the season, and expect to be told when they do.

Judged — a human decides these

These are the ones no emitter can settle: pacing, recovery and presence on air, whether a pitch landed, whether the room responded. The verdict is entered on the producer panel — which names where the call is recorded, not who forms it; an Executive Manager ruling lands on the same panel.

WhatEffectWhy a human and not a gate
A segment executed flawlessly on air
every seat
not setGenuinely human-judged
An aesthetic decision defended on air and upheld
this track
not setGenuinely human-judged, and the Creative Lead seat already requires willingness to have creative decisions reviewed on air (roles
Chat vibe check — the room responds to the segment
every seat
not setGenuinely human-judged
A pitch that fails its live review
every seat
not setGenuinely human-judged
Escalating without showing what you already tried
every seat
-15Human-judged: whether self-help was shown is a reading of a message, not a field
Manually doing repetitive work the established tooling could automate
every seat
-25Human-judged: 'could have been automated' is an engineering opinion about a specific task, and the Executive Manager's ruling is entered on the panel

The cap on a single week, and the rules that outrun it

Total adjustments in any one week are capped at capped (see the agreement) of that week's rate. On this track the first round pays — a week, so the cap is — — the tightest it ever gets, because the first round is the lowest rate on the ladder.

RuleEffectPast the cap by
Exposing an API key, token or secret in the repository-100— over
Late to an obligatory meeting, live broadcast or event-50— over
Pushing unapproved or un-QA'd work to a live production surface-50— over
3 rules on this seat exceed a whole week's cap alone. The cap still holds — no week can cost more than — — so the excess does not vanish, it means the rule is bounded by the cap rather than by its own number, and a second breach in the same week costs nothing extra. That is a real oddity in the schedule and it is printed rather than smoothed over: these are the Chairman's figures, reported not rewritten, and engine/gates/verify_scoring.py reports the same thing on every run.

Week one is exempt from all of it — see the pay section below — so the cap first bites in week two.

Not set9 of the rules above have no point value yet.
The RULE is real; the NUMBER has not been set by the Chairman, so it renders as not set rather than as a figure. governance/scoring.json forbids printing a placeholder on a candidate page, and engine/gates/verify_scoring.py lists every one of them on every run. You will not be scored against a number you were never shown.

The ones written for this track

Everything this seat needs to read

Conditions

Every seat, without exception. Every participant must have their own Kick, Twitch, YouTube, Instagram, TikTok and Facebook channels, all six available for joint co-streams. This is a condition of the seat, not a preference. Nobody appears on camera before counsel-reviewed release forms are signed. The Chairman may stop the programme at any time, with payment pro-rated by the day. How the record is read: The record is read from the system — registry entries executed, gates passed, work published. Where a rule genuinely cannot be read by a machine, governance/scoring.json marks it source:producer and says so on the page; nothing is scored as machine-observed that a human actually decided.

What this seat is paid

The rate rises as the field narrows — fewer people, more work, more pay. Read from governance/pay.json; this page holds no copy of it.

TrackRound 1
wks 1–4 · 3 seats
Round 2
wks 5–8 · 2 seats
Final
wks 9–13 · 1 seat
Creative Lead—/wk—/wk—/wk

What winning actually gets you

The stipend is the compensation. Surviving all 13 weeks on this track pays — — stated in the signed agreement depending on track, and less if you are eliminated early. Read the offer as that number and nothing else.

The post-season seat is not guaranteed. A post-season seat may be offered to any finalist at the Chairman's discretion. It is NOT guaranteed, and no candidate should accept on the expectation of one. It is said in these terms deliberately: a candidate who reads “the seat is the real prize” and then finds it discretionary was caught off guard, which is the one thing this framework forbids.

Week one, and the floor under every other week

Week one is paid in full regardless of elimination or adjustment. Week one is learning the machine. It removes the worst case — eliminated on Tuesday of week two, publicly, having earned very little — and it makes the offer defensible to a candidate reading the terms cold.

Total adjustments in any one week may not exceed capped (see the agreement) of that week's rate. Uncapped, the twelve rules could exceed the weekly rate — the spot-quiz alone at the agreed figure twice daily is the agreed figure against an EM's the agreed figure Round 1 week. A week worked for nothing is not a deterrent, it is an indefensible offer. On this track the worst possible Round 1 week still pays — — the rate less the cap, which is arithmetic rather than a promise.

The full adjustment schedule is on the operations policy, and every line of it has to be in front of you in writing before you sign.

All five roles · the show