All articles
I thought two skills shipped. Five weeks, invisible.
I shipped two skills that published green at 100 out of 100, moved the cards to Done, and only found out five weeks later that no customer could see either one. `tier` defaults to `internal`. That gap is why the Skill Creator profession exists — and why its eighth workflow refuses to call anything shipped until it has fetched the artifact and read it back.
One install path gave me 1 file. The other gave 166.
I installed the same skill twice, two documented ways, and counted the files. One gave me a single SKILL.md. The other gave me 166. The single-file version looked fine — it just quietly wasn't going to work. That finding is why @di-atomic/propagate v0.1.0 exists.
A press release is not a press department
Installing a press skill gives you a well-formed release. It does not give you the person who decides who receives it, under what embargo, with which boilerplate — or what to say when a journalist replies. That gap is the pr-manager profession, and its sharpest feature is what it refuses to do.
I checked 30 newsrooms for FOR IMMEDIATE RELEASE. None of them print it.
The most famous line in PR is gone from every modern newsroom I checked, and so is the ### end marker. Not because the press release died, but because its audience changed: answer engines now read it and cite it permanently, which moves the whole quality bar.
The Email Marketing Specialist: one AI role that runs outbound — and won't send from a cold domain
I stopped installing cold-email tools and hired one role instead. @di-atomic/email-outreach is an Email Marketing Specialist you equip onto an AI agent: it runs outbound and retention email end to end, gates deliverability before every send, and won't ship from a cold domain.
I shipped a profession that promised a guild. Then I shipped it empty.
Three days ago I shipped my first profession — and shipped its compounding guild empty. An OPVS review caught it. Here is what v0.2.0 put behind the address: three reference docs, four sealed learnings from real launch runs, a forkable registry seed, and why a seeded guild is the whole difference between a bundle and a profession.
I wrote up my best client win. Then legal killed it.
A year ago I wrote up my best client win, then asked for sign-off, and legal killed it — months of proof sunk one signature short. Here's the fix I built into a skill: get the release before the interview, source every number, and let the client hold the pen.
My audit found what was broken. Then everyone stared at the list.
Cutting a form from 11 fields to 4 lifted conversions 120%. Cutting one from 9 to 6 dropped them 14%. Both are real, which is why "reduce form fields" is not a rule. Here is the skill I built to turn an audit into a plan you can start on Monday, and why it refuses to write down 27%.
Stop asking if your numbers are good. Ask who's reading them.
Two founders showed me the same 26-month CAC payback. I told one it was fine and the other it was why they would not raise. Both reads were correct. Here is why a number carries a verdict only for someone, and the KPI skill I built to stop AI telling founders their numbers look solid.
Get new posts and release notes in your feed reader