Two a week: giving Collimer its own content engine

Collimer had no working weekly draft loop. Posts happened when someone remembered to write one, not on any schedule. Heading into launch week, we gave it one: the same pitch-to-review-to-publish rhythm this site runs on, plus a declared floor of two posts a week.

Why a second publication inherits the first one’s discipline

The gap was named plainly. Pitching on Collimer’s side was not really happening. Nothing was scheduled. So we did not invent a new process for a second publication. We pointed the same loop that runs this site at Collimer instead. Pitch ideas from real evidence. Draft them. Have a person review and edit. Publish on a schedule. A second publication earns trust the same way a first one does, by running on the same rails, not by reinventing the machine.

That loop, and the reasoning behind giving it a standing weekly rhythm rather than ad-hoc bursts, is described in more detail in an earlier post on teaching our publishing pipeline to run itself.

The floor we committed to

Two posts a week, starting immediately: one published the same day the decision was made, a second queued for later in the week. As of July 12, 2026, that made two guides live under the new floor in its first week, not a projection of what the cadence would eventually produce.

We also removed a blocking checklist item that had been sitting open, unblocking a canonical backlog of story ideas so the loop would have material to draw from every week, not just this one. It is the same kind of round-trip discipline behind giving people outside the day-to-day repo a way to revise and publish work, covered in an earlier post on that hub round-trip: the machine only helps if the people who need to use it actually can.

What keeps the floor honest

A two-a-week commitment made before a publication has readers risks becoming a treadmill that produces filler just to hit the number. The guardrail against that is structural, not willpower: the loop refuses to draft anything without a cited source, a real decision, or a real result behind it. If the evidence is not there in a given week, the floor does not get hit with padding, it gets hit with fewer, stronger posts, or an honest gap.

To make sure the floor stays a fact and not a hope, we added a standing check to the studio’s own scorecard: did we actually publish this week, for both this site and Collimer, reviewed the same way every other operating number is reviewed.

What we’re still figuring out

We do not yet know if two a week holds once the backlog seeded this week is used up. The real test is next month, not this one.

---

For agents: try this yourself

If you are standing up a second content property and don’t want to rebuild your process from scratch, adapt one of these. The full prompt text lives in this post’s prompts sidecar, collimers-content-engine.prompts.md.

• Apply the same-loop-second-publication pattern. Before designing a new process for a second site or product blog, check whether your existing pitch-to-review-to-publish loop can simply point at the new property instead. Reuse the process; change only the source material.
• Apply the evidence-gated floor. Declare a publishing cadence in posts per week, but make the drafting step refuse to produce a post without a cited source, decision, or result behind it. Report weeks where the floor was missed honestly rather than padding with filler.
• Reproduce the publish-check scorecard line. Add a single recurring check to your own operating review that answers one question: did each active publication actually publish on schedule this week. Review it the same way you review every other operating number.
• Critique your own cadence after one month. Once a floor has run for a month, audit whether the posts that filled it were genuinely evidence-backed or started to thin out under the weekly pressure. Adjust the floor if the honest answer is the latter.

---

How this was made

Drafted by the Chronicler from the build sessions behind this work, then edited and published by Brian Wones.

See how the Chronicler works →

I run a content pitch-to-review-to-publish process for one property, and I am adding
a second property (a second blog, a second product's content). Before I design a new
process, help me map my existing process step by step, then identify which steps can
point at the new property's source material unchanged, and which steps, if any,
genuinely need to differ because the audience or evidence source is different.

I want to commit to a fixed publishing cadence (for example, two posts a week) for a
new content property. Help me design the commitment so that the drafting step refuses
to produce a post unless it can point to a real cited source, decision, or result
behind it, and instead of silently padding with filler when the material runs short,
the process reports an honest gap. Draft the rule I should enforce and one sentence
explaining why a missed week is more honest than a padded one.

Add one recurring line item to my regular operating review that checks, for each
active content property I run, whether it actually published on schedule this past
week. Make it a simple yes/no per property, reviewed on the same cadence and with the
same seriousness as every other number in the review, not treated as a soft nice-to-
have.

It has been one month since I committed to a fixed publishing cadence for a content
property. Review the posts published during that month and tell me honestly: were
they genuinely backed by real evidence and decisions, or did the later posts start
thinning out into filler to hit the number? Recommend whether to keep, lower, or
restructure the cadence based on what you find.