A cadence you havenever timed is an intention.
Cadence does not collapse in week one. It collapses in week six, when the buffer runs out and the hours were always a guess. This grades the schedule you committed to against the hours you actually have — and refuses to grade it at all when the number underneath was never timed.
It does not collapse in week one. It collapses in week six.
In the shipped sample two commitments score exactly 92. One reads SUSTAINABLE and one reads ALREADY BROKEN, and the only difference between them is whether anyone ever timed the work.
Two commitments at 6.00 and 6.01 hours against 12 available both read SUSTAINABLE on their own. The plan reads ALREADY BROKEN. That override is arithmetic, not judgement.
Your available hours, your utilisation ceiling, your target buffer, your cadence. The only number we set is the sample-size floor for calling something measured, and the page states it plainly.
Nobody abandons a publishing schedule in week one. Week one is easy: the buffer is full, the enthusiasm is fresh, and the hours-per-unit figure is a number somebody said on a call. Week six is when the buffer hits zero, the real hours turn out to be double, and the schedule quietly stops being a schedule. All of that is visible in advance, from arithmetic you already have.
Change one number — units measured — and the verdict flips.
P-04 is loaded. It scores 92 and reads ALREADY BROKEN, because units measured is 0. Set it to 3 and the verdict flips — then look at the plan line, which does not.
| Commitment | Hrs/wk | Score | Evidence | Verdict |
|---|---|---|---|---|
| P-01 LinkedIn posts | 1.50 | 92 | MEASURED | SUSTAINABLE |
| P-02 Long-form blog posts | 4.00 | 55 | MEASURED | WILL SLIP |
| P-03 YouTube videos | 6.00 | 84 | RECALLED▼ | WILL SLIP |
| P-04 Weekly newsletter | 2.00 | 92 | ESTIMATED▼ | ALREADY BROKEN |
| P-05 Short-form video | 3.75 | 26 | MEASURED | ALREADY BROKEN |
Editing P-04 — Weekly newsletter
Written recovery plan
Survived one full cycle
Not dependent on one person
Score
92
band SUSTAINABLE
Evidence ceiling
ESTIMATED
capped
Buffer
4.00
weeks of cover
Commitment verdict
ALREADY BROKEN
Fix first: Measure hours per unit - the plan rests on a number nobody has timed
Hours required
17.25
of 12 available — 144%
On estimated hours
1
2 rows capped
Drop first
P-05 - Short-form video, 3.75 hours a week
DISPOSITIVE FIRED — the plan requires more hours than exist. This is arithmetic, not judgement: every row could read SUSTAINABLE and the plan would still be broken.
Plan verdict
PLAN ALREADY BROKEN
A cadence you have never timed is not a cadence. It is an intention.
Your numbers only — no benchmark, no industry average. It grades plans, never people, and connects to nothing. Not financial advice.
P-04 is loaded by default. It scores 92 — a strong, well-buffered, low-utilisation commitment — and reads ALREADY BROKEN, because units measured is 0. Set it to 3 and the verdict flips to SUSTAINABLE without a single other change. Then look at the plan line: it does not move, because 17.25 hours still do not fit in 12.
This is the real output of the shipped engine on the shipped sample.
Nothing below is written by hand. It is the verbatim stdout of python3 engine.py commitments-sample.csv, captured programmatically and diffed against this page before publishing.
==============================================================================
PUBLISHING CADENCE BACKSTOP - PCB-079 - RedHub AI LLC
==============================================================================
Your numbers only - no benchmark, no industry average.
Grades plans, never people. Nothing is fetched and nothing is sent.
Hours available per week 12.00
Your utilisation ceiling 60%
Your target buffer 2.00 weeks
------------------------------------------------------------------------------
COMMITMENT P-01 [LinkedIn posts]
------------------------------------------------------------------------------
Cadence 3.00 units per week
Hours per unit 0.50 (from 12 measured units -> MEASURED)
Hours required per week 1.50 (13% of available)
Buffer 8.00 units -> 2.67 weeks of cover
Utilisation headroom 2/2 weight 26 (derived)
Buffer depth 2/2 weight 22 (derived)
Written recovery plan 2/2 weight 18
Survived one full cycle 2/2 weight 18
Not dependent on one person 1/2 weight 16
Score 92/100 -> SUSTAINABLE
Evidence tier MEASURED -> ceiling SUSTAINABLE
Ceiling inert - the evidence supports the score
VERDICT SUSTAINABLE
Fix first Not dependent on one person
------------------------------------------------------------------------------
COMMITMENT P-02 [Long-form blog posts]
------------------------------------------------------------------------------
Cadence 1.00 units per week
Hours per unit 4.00 (from 5 measured units -> MEASURED)
Hours required per week 4.00 (33% of available)
Buffer 1.00 units -> 1.00 weeks of cover
Utilisation headroom 2/2 weight 26 (derived)
Buffer depth 1/2 weight 22 (derived)
Written recovery plan 1/2 weight 18
Survived one full cycle 1/2 weight 18
Not dependent on one person 0/2 weight 16
Score 55/100 -> WILL SLIP
Evidence tier MEASURED -> ceiling SUSTAINABLE
Ceiling inert - the evidence supports the score
VERDICT WILL SLIP
Fix first Not dependent on one person
------------------------------------------------------------------------------
COMMITMENT P-03 [YouTube videos]
------------------------------------------------------------------------------
Cadence 1.00 units per week
Hours per unit 6.00 (from 1 measured unit -> RECALLED)
Hours required per week 6.00 (50% of available)
Buffer 2.00 units -> 2.00 weeks of cover
Utilisation headroom 2/2 weight 26 (derived)
Buffer depth 2/2 weight 22 (derived)
Written recovery plan 2/2 weight 18
Survived one full cycle 2/2 weight 18
Not dependent on one person 0/2 weight 16
Score 84/100 -> SUSTAINABLE
Evidence tier RECALLED -> ceiling WILL SLIP
CEILING FIRED the verdict cannot be better than the evidence
under it. Capped at WILL SLIP.
VERDICT WILL SLIP
Fix first Measure hours per unit - the plan rests on a number nobody has timed
------------------------------------------------------------------------------
COMMITMENT P-04 [Weekly newsletter]
------------------------------------------------------------------------------
Cadence 1.00 units per week
Hours per unit 2.00 (from 0 measured units -> ESTIMATED)
Hours required per week 2.00 (17% of available)
Buffer 4.00 units -> 4.00 weeks of cover
Utilisation headroom 2/2 weight 26 (derived)
Buffer depth 2/2 weight 22 (derived)
Written recovery plan 2/2 weight 18
Survived one full cycle 2/2 weight 18
Not dependent on one person 1/2 weight 16
Score 92/100 -> SUSTAINABLE
Evidence tier ESTIMATED -> ceiling ALREADY BROKEN
CEILING FIRED the verdict cannot be better than the evidence
under it. Capped at ALREADY BROKEN.
VERDICT ALREADY BROKEN
Fix first Measure hours per unit - the plan rests on a number nobody has timed
------------------------------------------------------------------------------
COMMITMENT P-05 [Short-form video]
------------------------------------------------------------------------------
Cadence 5.00 units per week
Hours per unit 0.75 (from 20 measured units -> MEASURED)
Hours required per week 3.75 (31% of available)
Buffer 0.00 units -> 0.00 weeks of cover
Utilisation headroom 2/2 weight 26 (derived)
Buffer depth 0/2 weight 22 (derived)
Written recovery plan 0/2 weight 18
Survived one full cycle 0/2 weight 18
Not dependent on one person 0/2 weight 16
Score 26/100 -> ALREADY BROKEN
Evidence tier MEASURED -> ceiling SUSTAINABLE
Ceiling inert - the evidence supports the score
VERDICT ALREADY BROKEN
Fix first Buffer depth
==============================================================================
THE PLAN
==============================================================================
Commitments graded 5
Hours required, all rows 17.25
Hours available 12.00
Plan utilisation 144%
Rows on estimated hours 1
Rows capped by evidence 2
Worst row verdict ALREADY BROKEN
DISPOSITIVE FIRED the plan requires more hours than exist. This is
arithmetic, not judgement: every row could read
SUSTAINABLE and the plan would still be broken.
PLAN VERDICT PLAN ALREADY BROKEN
Drop first P-05 - Short-form video, 3.75 hours a week
==============================================================================Five weighted signals, and two things that sit above them.
Utilisation headroom 26, buffer depth 22, written recovery plan 18, survived one full cycle 18, not dependent on one person 16. Weights sum to 100. SUSTAINABLE at 75 or more, WILL SLIP from 50 to 74, ALREADY BROKEN under 50. The first two are derived from your own numbers, not marked — you cannot mark your way past whether the hours fit.
The verdict cannot be better than the evidence under it. Three or more measured units → MEASURED, no cap. One or two → RECALLED, capped at WILL SLIP. Zero → ESTIMATED, capped at ALREADY BROKEN. If hours-per-unit was never timed, every figure downstream inherits the guess, and an unknown reads as the worst state it could have been.
If the hours the whole plan requires exceed the hours you have, the plan reads PLAN ALREADY BROKEN however every row reads. Each commitment is graded against your hours; the plan is graded against those same hours once, for all of them together. This is the only override in the lane that can contradict every single row at once.
Three units is a stated methodological choice, not research and not an industry benchmark: one unit is an anecdote and two is a coincidence. It is the only number in the product we set rather than you, and the workbook says so on its first tab. Everything else — available hours, utilisation ceiling, target buffer, cadence — is yours.
It stops naming a signal and names the measurement instead. Every signal can be at 2 and the plan still rest on a number nobody has timed, so pointing you at a signal there would send you to improve something that is already perfect. The verifier asserts this on a commitment scoring 100.
It grades a plan from your numbers. It does not supply the numbers.
- A capacity check on a schedule you have already committed to.
- Deterministic and offline — same inputs, same verdict, every time.
- A named model you can argue with: the five weights, both bands, the tier thresholds and the dispositive are all on this page.
- Dateless. Re-run it next quarter with next quarter’s numbers.
- Not connected to anything — no calendar, no CMS, no scheduler. You bring the counts.
- Not a prediction. A SUSTAINABLE verdict means the arithmetic fits today, not that the cadence will hold.
- Not a productivity system. It will not make a unit take fewer hours; it will tell you how many you have.
- Not financial advice, and it scores no people.
Scope note: a SUSTAINABLE verdict means one plan, against hours you supplied, fits today. It is not a forecast and it guarantees no publishing, traffic or revenue outcome. The output is only as honest as the hours you enter — which is precisely why the evidence ceiling exists, and why a plan built on untimed work cannot reach the top verdict no matter how good the rest of it looks.
Anyone who has committed to a schedule they have not costed.
- Solo founders who committed to four channels in one enthusiastic afternoon.
- Anyone whose last content plan died quietly around week six.
- Small teams deciding what to drop before the quarter starts, not after.
- Creators publishing from an empty shelf every single week.
- Anyone wanting a forecast of whether the cadence holds. It measures fit, not fate.
- Teams with dedicated production staff and real capacity planning already.
- Making the work faster — that is a different problem.
- Evaluating the person doing the publishing. It grades the plan.
The Backstop tells you the hours do not fit. These three change the hours.
The direct lever on hours per unit. If the Backstop says the hours do not fit, this is the honest way to make a unit cost less of them.
ViewCadence pressure is where quality goes first. This is the guardrail for the weeks when you are publishing to hold the schedule.
ViewWhen the Backstop says drop first, this is what a single surviving commitment can look like when you give it the recovered hours.
ViewThe mechanics, stated plainly.
Week six is coming.
Do the arithmetic now.
One purchase, lifetime access, 12 months of updates. $79, once.
Deterministic and offline. Your numbers only. It grades plans never people, predicts nothing, and guarantees no publishing outcome. Not financial advice.
Sold by RedHub AI LLC · Secured by Stripe · redhub.ai