e12a063aa9
Pair 3 (internal target Thu 2026-11-12, six days early): - agent-worktree-contention — 'How do you manage worktrees when your agents end up waiting on each other?' (backlog 1xlz2bs, Scan 6 #4); git worktree + Docker Compose project-name docs fetched and read 2026-11-05. - agent-keeps-asking — 'Why does your coding agent keep stopping to ask you questions?' (handback cluster: 1xlfqmo, 1xk0ghj, 1xlxw0q); permission + hooks claims re-verified live 2026-11-05 against code.claude.com; the 1xlfqmo poster's measurement figures deliberately not cited. Pitch narrowed per CTO reopening verdict: trial→paid placeholder deleted (FRE-202 cohort too young; ratio waits for FRE-197), second breakage chosen (NOOK-815 uninvoked suite), any count named as activations, downloads never quoted.
5.0 KiB
Newsletter pitch — post-mortem shape (narrowed; no conversion figure)
Status: draft — narrowed 2026-11-06 per the CTO's reopening verdict on FRE-199. The "X% trial→paid" half of the shape cannot ship: the FRE-202 snapshot (window 2026-10-22 → 10-29) reads trial started 2, purchase 0, activation 0, with 291 of 291 analytics rows predating the FRE-366 instrumentation — a trial→paid ratio from that cohort is not a metric. The ratio comes from FRE-197's mature Day 28–35 cohort read when it lands. No figure in this file is a measurement. Send gates at the bottom.
Audience & channel
Editors of free dev-tool newsletters (the ones that run post-mortems and "what we learned" pieces). Direct email, one editor at a time, personalised. No paid placements at any budget without a CoS + sponsor decision.
The claim the pitch goes out with
A post-mortem pitch, not a product pitch: here is what shipping a Mac utility alone actually broke, and what we changed because of it. The product is the setting; the breakage and the fix are the story. The story stands on its own with zero funnel numbers — if no verified number exists on send day, the pitch ships without one rather than with a placeholder.
The pitch (shape — names and dates personalised; nothing else is a placeholder)
Subject: What shipping a Mac utility alone broke (post-mortem)
Hi ,
We build The Nook, a $10 Mac utility that surfaces coding-agent approvals at the notch. I'm not pitching the app — I'm offering a post-mortem your readers usually only get from funded startups, because shipping alone broke things in ways bigger teams don't hit:
- The silent update. In October we shipped a release whose appcast published cleanly — while every delta it named 404'd, so users on older builds were being offered updates that didn't exist. The publish path deleted files before the feed that named them was live. Nothing alerted; the feed just lied quietly. The fix — upload the feed's enclosures, verify every enclosure URL, copy the appcast up, and only then prune — is now ordered so the deletion can't run first.
- The green suite. Weeks earlier, a regression in our public relay shipped behind a passing test run — the suite that would have caught it existed but nothing invoked it. A green checkmark that covers nothing looks identical to one that covers everything. The suite is now wired into the standard test entry point, and the relay's deploy patcher is tested alongside it.
- What we're not going to quote you. A growth number, right now, would be theatre: our funnel is young and a conversion ratio from an immature cohort is not a measurement. If we attach a count, it will be activations — one per activated install, from our own relay store — and we will say exactly that, because download counts (a different number, from a different store) flatter nothing and confuse everyone. The receipts that matter here are the failure modes above, not a traction line.
The angle for your readers: solo distribution is a chain of quiet failure modes, each one invisible until it costs a user. We kept the receipts.
Happy to send the full write-up, screenshots of the broken feed, and the exact publish-order and test-wiring changes.
— , FrenoCorp
Where a number could come from, and what disqualifies it
- Trial→paid ratio: cannot ship today. FRE-202's snapshot shows a cohort of 2 trials, 0 purchases, pre-instrumentation rows. The ratio waits for FRE-197's mature Day 28–35 read. If it lands before send day and the CTO confirms the query, the ratio may be added — and it must say it is per install (install cap 3), not per person.
- Activations: the only count eligible to ship. One row per activated
install (
installs,created_at) in the relay store; CTO confirms the query before any send. The pitch names it as activations, never as downloads. - Downloads (S3 side) are never quoted. Different store, automation-heavy, and the pitch that says "downloads" when it means activations burns the credibility the post-mortem is built on.
- No number is estimated to fill any gap, ever.
Gates before anything is sent
- Any count quoted is activations, from a CTO-confirmed query — owner: CTO.
- The trial→paid ratio ships only from FRE-197's mature cohort read with CTO confirmation; otherwise the "what we're not going to quote you" paragraph stands as written — owner: CTO.
- Both breakage paragraphs are fact-checked against the engineering record (FRE-542 publish-order incident; the NOOK-815 uninvited-suite relay regression) before send — owner: CMO, evidence to CTO for verdict.
- Sponsor approves the send and holds the sending account (humans hold every account) — owner: Michael Freno.
Target dates (internal)
- Draft with placeholders: done 2026-11-05; narrowed to the no-figure shape 2026-11-06 (CTO direction).
- Send: after gates 1–4 — with the no-figure shape, the only hard input left is the CTO's query confirmation and the sponsor's send.