CONTENT-ENGINE.md
A posting system with a material pipeline, rather than a calendar that dies in week three. Content calendars fail because they schedule topics without material. This file schedules the material instead: where it comes from, who captures it, what the format rules are and what is off-position — so an assistant asked for a post has something real to work from.
# Content engine — <name>
## The claim
<One sentence. The thing this account is for. Everything published has to ladder into it.>
## The themes
| Theme | The angle it takes | Roughly what share of posts |
|---|---|---|
| <The problem> | <why the close is long, and why the usual explanation is wrong> | <40%> |
| <The method> | <how we actually do it, including what it costs> | <40%> |
| <The evidence> | <what we have seen across customers, with numbers> | <20%> |
## Where material comes from
This is the section that decides whether the account survives a busy month.
| Source | Who captures it | How |
|---|---|---|
| <Customer calls> | <whoever was on it> | <one line in the #material channel within an hour> |
| <Support threads> | <support> | <weekly, anything asked twice> |
| <Our own analysis> | <founder> | <whenever a number surprises us> |
| <Things we got wrong> | <anyone> | <same day, while it still stings> |
Rule: no post without material. A week with no material is a week with no posts, and that is the correct outcome.
## Format rules
- 120–220 words.
- The first line states the claim. No teasing.
- One concrete detail with real numbers.
- One line naming the limitation of the claim.
- Ends with something answerable.
- Maximum two hashtags, only if genuinely searched.
- No hook formulas, no emoji bullets, no engagement bait, no one-word-per-line ladders.
See VOICE.md for the voice; it overrides anything here on style.
## The not-list
<Topics that dilute the position, including the ones that perform well:
- General AI commentary
- Hiring and culture posts
- Industry news reactions with no angle of our own
- Anything about "the future of finance">
## Cadence
<Two posts a week, Tuesday and Thursday morning. Comments daily, ten minutes, on the accounts listed below. Comments matter more than posts for the first three months because we have no reach yet.>
## Accounts to engage with
<A list of 15–25 accounts whose audience is our audience. Comment before posting, not after.>
## The belonging test
<Before publishing: does this make the claim more credible, or is it just true? "Just true" is not a reason to post.>
## Review
Monthly: which posts produced conversations, not which produced likes. Likes and pipeline correlate poorly enough that tracking the first one actively misleads.- Claude Project knowledge
- ChatGPT Project files
- The claim and themes (from the positioning brief), and the material sources
Everything in <angle brackets> is example content, not instructions. Replace it. The file ships filled in rather than empty on purpose: an empty template gets abandoned, a filled one gets edited.
The reference behind this file
Every post on this site starts from a number we checked. Useful as a worked example of "no post without material".
Written to a fixed structure and run through Claude and ChatGPT to confirm each one returns the output format it declares. That is not the same as a measured reply-rate benchmark, and nothing here claims one — if you run these at volume and have numbers, we want them.
Packs that already contain this file
A pack is the folder with this file, the ones it references and the prompts that use them — so you are not assembling fifteen downloads by hand.
One file is a start. The folder is the point.
Files reference each other — CLAUDE.md declares the reading order, the prompts expect the files by name. The packs ship them together, already wired up.