System Prompts Leaks is a collection of captured system instructions used to guide AI chatbots and coding agents before they receive user messages. It serves researchers and developers studying how different AI assistants are directed.
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 asgeirtj/system_prompts_leaks --skill make-a-docgit clone --depth 1 https://github.com/asgeirtj/system_prompts_leaksWrote 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/asgeirtj/system_prompts_leaks/make-a-doc)<a href="https://agentmods.dev/skills/asgeirtj/system_prompts_leaks/make-a-doc"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/make-a-doc/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/asgeirtj/system_prompts_leaks/make-a-doc"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/make-a-doc.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Snyk pass
- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high System Prompt Leakage · line 24 Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.Fix: Remove any instructions that reveal, print, or output system prompts or internal rules. System instructions should never be exposed to end users.
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 | $0.00013 | $0.00655 |
| Opus 5 | $0.00006 | $0.00328 |
| Sonnet 5 | $0.00003 | $0.00131 |
| Haiku 4.5 | $0.00001 | $0.00065 |
Grade A, and why
make-a-doc 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 11d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- make-a-doc — 92% identical, 3 lines differ
What it actually says
Make a doc
Create a document (resume, one-pager, memo, letter, report, guide, paper). First decide which of two shapes the user wants — they export completely differently:
Flowing pages — text that pours onto standard sheets (Letter/A4)
and breaks wherever needed: reports, memos, letters, papers, guides.
START by calling copy_starter_component with
kind: "doc_page.js", then write the whole document as normal
flowing HTML inside <doc-page size="letter" margin="0.75in">.
The component owns the sheet, the desk background, and all print
geometry — do NOT write your own @page rule, body background,
page-card divs, break-after: page fake sheets, or
break-inside: avoid on items inside multi-column grids (a grid only
breaks between rows, so a kept row that doesn't fit leaves a blank band).
Print rules for flowing pages: multi-column text uses CSS columns
(column-count + column-gap; column-span: all on a heading
that spans; hyphens: auto in narrow columns — it needs lang
on the html element), never side-by-side
flex/grid columns — only real CSS columns flow and break across pages.
Use break-before: page on anything that must start a new page (a
chapter, an appendix); add custom kept-together blocks (callouts, stat
tiles, cards) to a break-inside: avoid rule and keep each shorter
than a page — the component already keeps headings with their content,
keeps figures/code/table rows whole, and suppresses orphans and widows
(extend orphans: 3; widows: 3 to custom text blocks). Long tables
get a <thead> so the header repeats on every page. No
position: fixed/sticky and no viewport units in content —
fixed elements stamp every printed page (running headers/footers go in
the component's slots) and 100vh mis-sizes at print.
Fixed sheet — a design that must fill exactly one page of fixed
dimensions: poster, infographic, social graphic, certificate. No
starter component — build it at its true pixel size with an explicit
px width (and height if fixed) on the top-level element; the
export sizes the PDF page to it automatically. Do not write any
@page rule for it.
Styling (both shapes): body type 14–16px with generous line-height (1.55–1.7); clear heading hierarchy; restrained palette. Tables get a header row and hairline borders; figures and code blocks each carry a short caption. Open with the document's own h1 as the first body element (use any header-shaped first line of pasted content as that h1 rather than rendering it as a separate masthead).
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.
- 11d ago First seen · 53 lines · 13 tokens per session scan A 967825a87932
make-a-doc is a skill published in the GitHub repository asgeirtj/system_prompts_leaks (64,614 stars, last pushed today), licensed CC0-1.0. It adds 13 tokens to every session and 655 once invoked, about $0.0001 per session on Opus 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 2026-08-30.
Other skills, from other repositories
dcf-model
Build discounted cash flow valuation workbooks in Excel.
comps-analysis
Build comparable-company valuation workbooks in Excel.
lbo-model
Build leveraged buyout workbooks with IRR/MOIC in Excel.
google-workspace
Gmail, Calendar, Drive, Docs, Sheets via gws CLI or Python.
PDF files: create, read, merge, fill, OCR, edit text.
excel-author
Build auditable financial workbooks headless via openpyxl.