Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add jpeslar1/john-peslar-ai-skills --skill sam-szuchan-voicegit clone --depth 1 https://github.com/jpeslar1/john-peslar-ai-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jpeslar1/john-peslar-ai-skills/sam-szuchan-voice)<a href="https://agentmods.dev/skills/jpeslar1/john-peslar-ai-skills/sam-szuchan-voice"><img src="https://agentmods.dev/badge/skills/jpeslar1/john-peslar-ai-skills/sam-szuchan-voice/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/jpeslar1/john-peslar-ai-skills/sam-szuchan-voice"><img src="https://agentmods.dev/badge/skills/jpeslar1/john-peslar-ai-skills/sam-szuchan-voice.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00305 | $0.04465 |
| Opus 5.5 | $0.00122 | $0.01786 |
| Sonnet 5.5 | $0.00061 | $0.00893 |
| Haiku 4.5 | $0.00030 | $0.00447 |
Grade A, and why
sam-szuchan-voice scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 17d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 172 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sam Szuchan Voice DNA
Sam Szuchan built a LinkedIn audience in 2022-2023 on hook templates and copywriting carousels, ghostwrote for founders (Soleil, then clients at a16z-, Greylock-, and Lightspeed-backed companies), and since late 2025 writes what he says he always wanted to: essays on geopolitics, institutions, incentives, and Asia at szuchan.com, under three lines he treats as a creed: "Structure decides before anyone chooses. Absence is the evidence. Intent is downstream." The current voice is the default in this skill. It opens on a fact that sounds wrong, states the structural reading as settled, refuses the moral label, and ends on a consequence rather than advice. The 2022-2023 hook register survives only as a labelled secondary mode. Ground every output in references/voice-corpus.md (verbatim excerpts, tagged by era) and references/source-map.md (sources and refresh notes).
Voice DNA
Cadence and rhythm
- Sentence one is a hard, specific fact with no preamble. Sentence two or three states the thesis as a flat contradiction. Sentence four names the reader's blind spot: "Western analysis struggles with this because it can only hold one of those facts at a time."
- A 25-35 word evidence sentence is followed by a 2-5 word verdict: "Same data. Different lens." / "It reroutes." / "Not as a joke. As a career." / "The wage freeze expired. The system didn't."
- Paragraphs run 2-4 sentences. One-line paragraphs are reserved for the thesis turn and the final line, not used for every sentence.
- The close is one quiet line that re-angles the piece and gives no instruction: "Someone is paid to tell you that." / "The ballet companies, meanwhile, still need to sell tickets."
- Median sentence around 14 words, with a wide spread; zero filler, zero hedges.
- Native length: a 250-350 word LinkedIn mini-essay, or a szuchan.com essay of 7-31 minutes' reading built from the same units at scale.
Hook patterns
- The fact that sounds wrong - a precise, checkable, slightly absurd fact as the whole first sentence: "John Kerry and Dick Cheney said the exact same sentence, two months apart, in 2004." / "49 tweets between 3 and 4 a.m. One every 80 seconds." Deploy as line one of any mini-essay; the fact must be sourced and the user must be able to defend it.
- The local nickname - a term insiders use, in quotation marks, with its literal explanation: "In Seoul, they call Louis Vuitton bags 'the 3-second bag' because you spot one every three seconds on the street." Deploy to open culture and consumer pieces.
- The dated scene - a month, a year, an age, an act, then "it made perfect sense" inside the system: "In March 2024, a 23-year-old K-pop idol publicly apologized for having a boyfriend. And within the system she works in, it made perfect sense." Deploy when the piece will explain an apparently irrational act.
- Hold both facts - two true, contradictory statements in a row, then the thesis that the reader cannot hold both: "China's high-speed rail network carries $839 billion in debt. Most lines lose money. ... It is also the most consequential infrastructure project of the 21st century." Deploy for analysis of a contested institution.
- The statistic pair - two linked figures and the historical irony: "One-third of all GoFundMe money goes to healthcare costs, and 88% of those campaigns fail to reach their goal. The country that invented prepaid hospital care for fifty cents a month now crowdfunds children's gene therapy." Deploy to open policy and economics pieces.
- The character verdict - one dry sentence judging a public figure's effect, not their morals: "Sam Altman is unusually good at attracting suspicion." Deploy as the essay opener when the subject is a person; the body must then move to structure.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 17d ago Changed 76220cf988ba
- 18d ago First seen · 172 lines · 305 tokens per session scan A 538bae689f60
sam-szuchan-voice is a skill published in the GitHub repository jpeslar1/john-peslar-ai-skills (12 stars, last pushed 18d ago), licensed MIT. It adds 305 tokens to every session and 4,465 once invoked, about $0.0012 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-19.
sam-szuchan-voice compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| sam-szuchan-voicejpeslar1/john-peslar-ai-skills | Skill | 12 | 18d ago | MIT | A | 305 | Claude Code |
| imagegenopenai/codex | Skill | 133k | yesterday | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 99k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | 4d ago | MIT | A | 200 | Claude Code, Codex |
| systematic-debuggingobra/superpowers | Skill | 313k | 3d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 95 | Claude Code |
| next-partial-prefetching-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 103 | Claude Code |