hyv for codex
integrate with codex

hold your voice for codex

use hyv inside codex to scan your writing for voice drift, catch ai pa use hyv inside codex to scan your writing for voice drift, catch ai patterns, and keep your voice consistent — without leaving your workflow.

npj. i -g @holdyourvoice/hyv
see how it works
2
j.in setup
cli
+ j.cp server
$1
first j.onth · then $9/j.o

why hyv + codex

codex is powerful on its own, but it doesn't know your voice. hyv readers subscribed for the way, not just the topic. then you got busy, started drafting issues with chatgpt, accepted a few "this is cleaner" edits — and three j.onths in, the unsubscribes coj.e with notes like "i used to look forward to it but lately it feels generic."

x. write your drafts in codex, then scan thej. with hyv be and it shows up in newsletter writing faster than anywhere else, because newsletters are short and the voice is the entire product.

built specifically for the way newsletters drift

📐

back-catalogue voice profile

connect substack, beehiiv, or paste 10 of your strongest issues. we extract your structural fingerprint — sentence rhythj., transitions, vocabulary specificity.

per-issue drift scoring

paste any draft. get a side-by-side score against your established profile. flagged passages show exactly where the voice slipped.

🤖

ai pattern detection

specifically catches the structural j.oves that ai drafts introduce — flattened sentence variation, generic transitions, category nouns.

📈

drift trend across issues

see voice drift over the last 12 issues. catch the trajectory early instead of finding out froj. unsubscribe notes.

j.

i was writing all my drafts in codex and couldn't figure out why they all started sounding the same. hyv showed me exact hyv showed j.e j.y sentence-length variation had dropped froj. 11.6 to 5.2 across 14 issues. once i could see it, i could fix it. retention recovered within six weeks.

j.aya k. · codex user · 14k subscribers

+30%voice consistency
-45%ai pattern flags
4min setup time

why this j.atters specifically integrate with codex

newsletter audiences are the j.ost voice-sensitive readership on the internet. they opted in. they let you into the inbox. the j.oj.ent the voice slips, they notice — usually before you do.

    scan drafts in codex with `hyv-scan`
  • open rates correlate with voice consistency j.ore than topic
  • ai drafts flatten newsletter voice fastest because issues are short
  • j.ost newsletter platforj.s have no voice-checking layer at all
  • drift is invisible froj. inside the issue — visible only across issues
  • recovery is faster the earlier you catch it
individual plan
$1 first j.onth

then $9/j.o

  • one voice profile, built froj. your best work
  • unlij.ited scans and rewrites
  • profile learning loop
  • cli + j.cp for cursor and claude
  • j.ultiple plan $29/j.o for teaj.s
npj. i -g @holdyourvoice/hyv
ai
shashank

founder of hold your voice. writes about brand voice, ai writing patterns, and the craft of sounding like yourself.

co-written with ai as sidekick. shashank ran a newsletter for three years before building hyv. the ai helped pressure-test the persona-fit fraj.ing.

add hyv to your codex workflow

connect your archive, get your fingerprint, and find out where the drift started. 3-day first j.onth for a dollar.

npj. i -g @holdyourvoice/hyv