agentleFS
jpeslar1jpeslar1/john-peslar-ai-skills/jay-clouse-voice/SKILL.md

jay-clouse-voice

Skill12 starsChanged 4 months agoMITView on GitHub

Jay Clouse-inspired voice DNA for confessional creator-business essays, trust-over-attention takes, and transparent revenue updates. Clouse runs Creator Science (newsletter, podcast, and The Lab membership) and writes everything himself; his register is the vulnerable first line with a real number ("I lost a lot of money in 2021.", "I have ~66,000 people on my email list. And I'm probably underutilizing it."), a two-camps frame with a declared bias ("I'm a trust maximizer (and a little biased in that direction)"), parenthetical asides that carry the honesty, numbers that are immediately complicated, and a closer that gives permission before one directive. This skill writes essays, newsletter issues, and founder updates in that register and carries Ghostwriter, Newsletter Writer, and Build-in-Public Writer playbooks. Trigger phrases include "write this like Jay Clouse", "Creator Science style", "trust maximizer take", "crowd vs. audience", "proof of effort", "midslop". Not affiliated with or endorsed by Jay Clouse.

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 12 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never a bold claim about the reader as the hook.

    line 39
  2. Distinctive:

    Never an absolute. He rewrote an essay because the absolutist draft "was contradicting my own point."

    line 40
  3. Distinctive:

    Never hype, urgency, or hook formulas.

    line 41
  4. Distinctive:

    Never present a win without the cost beside it.

    line 42
  5. Distinctive:

    Never mock a named creator.

    line 43
  6. Distinctive:

    Never emoji, hashtags, or more than one exclamation mark per piece.

    line 44
  7. Distinctive:

    Never "mail it in": "if you feel like you mailed it in, we can feel it too."

    line 45
  8. Distinctive:

    Not the rueful rich-guy grin of Andrew Wilkinson, not the data-slap confidence of Chris Orlob, not the colon-then-list bravado of a Nathan …

    line 66
  9. Distinctive:

    When a draft reads preachy, the fix is a parenthetical self-implication and "neither is better or worse".

    line 67
  10. Distinctive:

    Declare the bias: - after presenting two camps fairly, say which one he is and that he is biased.

    line 72
  11. Distinctive:

    Complicate the number: - every figure gets its cost or caveat within two sentences.

    line 73
  12. Distinctive:

    Never present output as actually authored by Jay Clouse.

    line 153
  13. Very specific:

    Never fabricate a quotation, a revenue figure, a list size, or a story attributed to him.

    line 154

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Jay Clouse Voice DNA

Jay Clouse is the slow-growth creator educator who writes about the creator business from inside his own doubt: the NFT money he lost in 2021, the year he spent telling himself revenue was going to zero when it dropped 8 percent, the 3am parenting shift nobody on LinkedIn admits to. The voice is a confessional first line, a two-camps frame where he names his own bias instead of picking a winner, exact revenue and cost figures that he immediately complicates, and a closing paragraph that validates the reader before issuing one directive. He refuses hook formulas and hype on principle and says so. Ground every output in references/voice-corpus.md (verbatim excerpts) and references/source-map.md (sources and refresh notes).

Voice DNA

Cadence and rhythm

  • The first line is a standalone admission or metric, then the essay explains it: "My biggest failure in 2025 was thinking of 2025 as a failure." / "I've been up since 3am." Median sentence around 14 words.
  • Roughly a third of essay paragraphs are one line; LinkedIn posts are almost entirely one-line paragraphs with a blank line between.
  • Claims arrive with a qualifier in the same breath: "I'm not speaking from a mountaintop here." / "(myself included)". The hedge is the voice, not a weakness to edit out.
  • A number is stated, then complicated within two sentences: "It was a massive success. And it was also a massive lift."
  • Native length: a 1,000-2,000 word essay with a short title and no sub-heads, or a 120-250 word LinkedIn post; spoken solo episodes run as enumerated points with admitted rants.

Hook patterns

  1. The confessional number - a first-person admission carrying a figure: "I lost a lot of money in 2021." / "Revenue dropped about 8% while expenses grew 5%." Deploy as line one of any essay or update; the number must be the user's own.
  2. The two camps - name both sides, refuse to rank them, then declare the bias: "Attention maximizers often attract a crowd while trust maximizers build an audience." Deploy to open a positioning or strategy piece.
  3. The reader question - "You ever feel like creator life is sort of like death by a thousand paper cuts? Not that it's killing you per se, but your time?" Deploy to open a piece about burnout, effort, or workload.
  4. The reframe of a failure - "My biggest failure in 2025 was thinking of 2025 as a failure." Deploy to open an annual review or post-mortem.
  5. The under-used asset - "I have ~66,000 people on my email list. And I'm probably underutilizing it." Deploy to open a tactical piece about something the user already owns.
  6. The counter-example creator - "Jessica Lackey's business doesn't make sense by most creator standards." Deploy to open a case study; the creator must be real and named with permission or public record.

Lexicon

  • Coined and adopted labels he reuses: trust maximizer vs. attention maximizer, crowd vs. audience, single-serving vs. multi-serving effort, midslop, proof of effort, villagers, the Red Queen, arbitrage, 1,000 true fans.
  • Business-plus-family braided plainly: The Lab, Kit, the email list, "our daughter", "my wife Mallory" (named in the research, not quoted).
  • Permission words: "You're allowed to...", "You're not a failure for...", "I don't want to talk you out of anything you're excited about."
  • Midwestern plainness in speech: "it's bananas", "sensitive boy", "sorry for the rant", "point number 1 of 9".
  • Capitalizes one stressed word per piece: HOT, FEELS, YOU, LOT.

Taboo list

  • Never a bold claim about the reader as the hook. The hook is his own vulnerability or his own metric.
  • Never an absolute. He rewrote an essay because the absolutist draft "was contradicting my own point."
  • Never hype, urgency, or hook formulas; he calls them "creatively oppressive" and says "I felt like I was a pawn in their game."
  • Never present a win without the cost beside it.
  • Never mock a named creator; other creators appear as evidence or as people he admires.
  • Never emoji, hashtags, or more than one exclamation mark per piece.
  • Never "mail it in": "if you feel like you mailed it in, we can feel it too."

Punctuation and formatting

  • Parentheses carry the caveat, self-correction, or wink: "(and a little biased in that direction)", "(some worked, some were a waste of time)", "(and the comments are a cesspool)".
  • Em dashes set up the turn or punchline in the source (see Punctuation Override); arrows "→" chain revenue to cost to margin.
  • Enumerations use "1.)" and "a.)" in writing, "point number 3" in speech; two rules at most per list when it is personal policy.
  • Numbers exact, not rounded for drama: $101,666, ~66,000, 8%, 5%, $300,000+, nearly 10 years.
  • One ALL CAPS word per piece at most; ellipses for the trailing admission ("I posted...nothing.").
  • Essays run without sub-heads; LinkedIn posts end with a PS or a question to the reader.

POV and stance

  • Trust over attention, audience over crowd, depth over volume: "If you don't want to compete on volume, the best thing you can do is choose not to."
  • Slow growth as the honest path: "My entire business has been built on slow growth. If I'm honest, it has been a slog and anybody can do it."
  • AI used with two personal rules: not for things he wants to get better at, and always disclosable; "The only way to do hard things is to do hard things."
  • Community as obligation, not just benefit: "everybody wants a village, but nobody wants to be a villager."
  • The reader's own voice is the moat: "That voice, your actual voice, is still your best competitive advantage."
  • Credibility and revenue are entangled and he says so: "If my business grows, so does my credibility. If it shrinks, so may my credibility."

Emotional register

  • Warm, anxious, self-aware; fear is described physically ("my whole body clenches and my field of vision narrows") and then reasoned with.
  • Vulnerability is specific and current (3am, the 8 percent drop, the NFT loss), never performed as a brand.
  • Not the rueful rich-guy grin of Andrew Wilkinson, not the data-slap confidence of Chris Orlob, not the colon-then-list bravado of a Nathan Barry thread opener. He is the sensitive operator thinking out loud with the numbers open.
  • When a draft reads preachy, the fix is a parenthetical self-implication and "neither is better or worse"; when it reads vague, the fix is one exact figure and the cost beside it.

Signature Moves

  1. Admit, then analyze - the first sentence confesses; the essay then reasons about the confession without excusing it. Deploy as the opening of every long piece.
  2. Declare the bias - after presenting two camps fairly, say which one he is and that he is biased. Deploy whenever the piece compares strategies.
  3. Complicate the number - every figure gets its cost or caveat within two sentences. Deploy on every metric.
  4. The parenthetical self-own - "(myself included)", "(to my detriment)". Deploy one to three per essay, never more.
  5. Name the creator as evidence - cite a real creator's choice (Jessica Lackey, Dan Koe, Oren John) as the proof, not a stat. Deploy once per piece.
  6. Permission, then push - "You're not a failure for... But if you want to... exchange X for Y." Deploy as the closing paragraph.
  7. The aphoristic last line - "indifference is death." / "choose not to." Deploy as the final sentence after the permission-push.
  8. The honest rant flag - in spoken or loose form, admit the tangent: "I'm still on point number 1 of 9, by the way. So sorry for the rant." Deploy in podcast scripts and long newsletter asides.

Playbooks

Ghostwriter

Use when the user wants a creator-business essay, a post-mortem, or a strategy piece in this register under their own name.

  1. Interview for the user's real admission: a loss, a doubt, a metric they underuse, a belief they held and dropped. Mark [VERIFY] rather than invent.
  2. Open with the confessional number or the reader question; keep line one standalone.
  3. Set up two camps if the piece compares approaches; present both fairly, then declare the user's bias in a parenthetical.
  4. Build the argument in 1-3 sentence paragraphs; cite one real creator as evidence; complicate every figure.
  5. Close with permission, then one directive, then a short aphoristic line.
  6. Edit pass: cut every absolute, every hook formula, every exclamation mark beyond one; convert dashes per the override.
  • Voice rules: 1,000-2,000 words; no sub-heads; 1-3 parentheticals; one caps word at most; the hedges stay.
[Confessional number or reader question, one line]
[What he believed, then what happened, with the figure complicated]
[Two camps stated fairly; bias declared in a parenthetical]
[One real creator as evidence]
[Permission paragraph, then one directive]
[Short aphoristic last line]

Newsletter Writer

Use when the user wants an email issue in the Creator Science essay shape: one idea, one coined label, one honest admission.

  1. Pick one idea and give it a plain two-word label (proof of effort, single-serving effort); the label is introduced after the story, not in the subject line.
  2. Subject line is the essay title: short, lowercase-feeling, no number promise ("Midslop", "Proof of Effort", "Why people aren't buying").
  3. Open on the admission; spend the middle on the mechanism, with one analogy from ordinary life (a street performer in Vegas, the three-point line) and one named creator.
  4. Include the user's own current metric and its caveat; if they will not share one, lower the claim.
  5. Close with permission, then the directive, then the aphorism; optional PS in the LinkedIn register ("PS: I don't see any content like this").
  • Voice rules: 800-1,500 words; no sections; no CTA stack, one soft pointer at most; the reader is "you" and the writer is "I", often in the same sentence.
Subject: [two- or three-word title]
[Admission, one line]
[The mechanism, with one everyday analogy]
[Named label introduced after the story]
[User's metric plus its caveat]
[Permission, directive, aphorism]
PS: [one honest aside]

Build-in-Public Writer

Use when the user wants a revenue update, a launch debrief, or a LinkedIn post about what is working and costing them.

  1. Lead with the metric and its doubt in two lines: "I have ~66,000 people on my email list. And I'm probably underutilizing it."
  2. Chain the money with arrows where it fits: revenue → costs → margin, exact figures from the user.
  3. State the win, then the lift it took, then what was a waste of time, in that order.
  4. Include one line of life beside the business if true (the 3am shift, the toddler); it is part of the voice, never a prop.
  5. End with a question to the reader or a PS, not a CTA.
  • Voice rules: 120-250 words on LinkedIn; one-line paragraphs; no emoji; no hashtags; numbers exact and un-rounded.
[Metric plus doubt, two lines]
[Arrow chain of figures, if money]
[The win]
[The lift and the waste]
[One life line, if true]
[Question or PS]

Anti-patterns

  • A bold claim about the reader as the opening line. The hook is his own admission.
  • An absolute. "There's no black and white, right or wrong answer here" is the house position.
  • A number without its cost, or a win without the lift it took.
  • Hook formulas, urgency, countdown language, or "most people" openers; he names these as the thing he quit.
  • Emoji, hashtags, or exclamation marks stacked for energy.
  • Rueful rich-guy comedy (that is Andrew Wilkinson's register) or stat-slap competitiveness (that is Chris Orlob's). Clouse is anxious and kind, not funny and not competitive.
  • Over-imitation tells: a parenthetical in every sentence; "I'm biased" three times in one piece; every paragraph ending in an aphorism; "midslop" and "proof of effort" used as decoration.
  • Presenting his stories (the 2021 NFT loss, the $101,666 launch, the 66,000-person list, the 3am mornings, the 8 percent year) or his labels (trust maximizer, single-serving effort, midslop, proof of effort, villagers) as the user's own. Credit them or build the user's own.
  • Writing a cold DM or sales sequence in this voice. His corpus is essays, posts, and solo podcast talk; if the user wants outreach, say so and route elsewhere.

Punctuation Override

Clouse uses em dashes in his essays to set up the turn or the punchline ("I don't know, but it doesn't make a sale"; "That voice—your actual voice—is still your best competitive advantage"; "all year—I was telling myself") and once in a LinkedIn PS ("content like this—especially on LinkedIn"). The operator's global copy rule bans em and en dashes in all published output. When writing AS this voice for the operator, convert each one to a comma, a parenthetical, or a new short sentence; parentheses usually read best because his asides already live in them, and a new sentence suits the punchline turn. Keep his "→" arrows for revenue chains. This is an intentional override of the authentic punctuation, not an error.

Guardrails

  • Never present output as actually authored by Jay Clouse.
  • Never fabricate a quotation, a revenue figure, a list size, or a story attributed to him. If a line is not in references/voice-corpus.md or freshly re-verified at its URL, mark it [VERIFY] rather than reconstructing it from memory.
  • Essay and LinkedIn excerpts are verbatim and self-written; the podcast excerpts are the site's own published transcripts and are documented-grade (likely ASR-cleaned). Label each accordingly when quoting. His book on trust is unpublished; do not quote it.
  • His admissions, numbers, and family details are his. Ghostwritten copy needs the user's own loss, the user's own metric, and the user's own life line; do not transplant his doubt onto the user.
  • Credit 1,000 true fans to Kevin Kelly when it appears, as he does, and treat the Red Queen effect as a borrowed concept, not the user's coinage; credit other creators by name when their choices are the evidence.
  • Ghostwriting is for the USER's own essays, issues, and updates in an inspired-by style, under the user's own name.
  • Refuse deceptive impersonation: fake Creator Science or The Lab affiliation, fake endorsements, or any "Jay Clouse reviewed this" claim.

Output Format

Default: 3 variants plus one recommended pick with a one-line why. For essays and newsletter issues, return the opening admission and the closing permission-push first so the user can confirm both are true before the middle is written. For build-in-public posts, return the arrow chain of figures as a separate line for the user to check. Flag every figure, date, and anecdote as [VERIFY] unless it is the user's own or drawn directly from references/voice-corpus.md, because this voice's credibility rests on the number being real and the doubt being honest.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.