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 save-as-pdfgit 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/save-as-pdf)<a href="https://agentmods.dev/skills/asgeirtj/system_prompts_leaks/save-as-pdf"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/save-as-pdf/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/save-as-pdf"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/save-as-pdf.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 25 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 | $0.00009 | $0.02067 |
| Opus 5 | $0.00005 | $0.01033 |
| Sonnet 5 | $0.00002 | $0.00413 |
| Haiku 4.5 | $0.00001 | $0.00207 |
Grade A, and why
save-as-pdf 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 12d 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 — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Save as PDF
Reformat the current HTML design into a paginated, paper-ready PDF. The "Instant" export already gives the user a PDF at the design's native pixel size — this path is for when they want real pages.
Do NOT rasterize the page into a PDF. Never use jsPDF, html2canvas, dom-to-image, or any other canvas/screenshot-to-PDF approach — they produce blurry, non-selectable, oversized output, and do not generate a PDF binary yourself. PDF export is print-based: a print-ready copy is handed to show_pdf_export_dialog and the browser's own print engine renders crisp, selectable, text-based pages. The only supported way to make a copy print-ready is a component that owns its print geometry — the doc_page starter for documents, or a source already built on <deck-stage> or <doc-page>. Do not hand-author @page rules or print CSS resets, and NEVER declare <meta name="omelette-owns-print"> yourself — the starters announce print ownership at runtime, and a hand-authored meta tells the platform to trust print CSS you would then have to hand-write and maintain. Hand-rolled print CSS behind that meta is the legacy path; the doc_page starter replaces it everywhere it is available.
Steps
-
Read the current HTML design file to understand its structure and content. Re-read it on every PDF request, even if you read it or made a print copy earlier in this conversation — the user may have changed content or tweak values (the Tweaks panel writes into the source file) since then. Note the
[version: v<N>]token in the read result's header — step 2's provenance stamp needs it, and only a version read in THIS request is valid to stamp. -
Write the print copy, stamped with its provenance. Always write it fresh from the source you just read — an existing
-printcopy from an earlier request is a stale snapshot, and reusing or only partially updating it ships outdated values to the PDF. The print file path is the source path with-printinserted before the extension — same directory, same basename. If the source isslides/deck.html, writeslides/deck-print.html; if the source isweb/index.html, writeweb/index-print.html. Do NOT use the deck title or project name as the filename, and do NOT write to the project root if the source is in a subdirectory — any change in directory depth breaks every relative URL (@font-facesrc: url(...),<img src>,<link href>, CSSbackground: url(...)) and the print tab shows missing images and system-font fallbacks.
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.
- 12d ago First seen · 54 lines · 9 tokens per session scan A 2ee629104dc8
save-as-pdf is a skill published in the GitHub repository asgeirtj/system_prompts_leaks (64,614 stars, last pushed 2d ago), licensed CC0-1.0. It adds 9 tokens to every session and 2,067 once invoked, about $0.0000 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
PDF files: create, read, merge, fill, OCR, edit text.
pptx
Create and validate Microsoft PowerPoint presentations (.pptx), including structured slide decks, tables, workflows, metadata, and reproducible generation scripts. Use for presentation, slides, PowerPoint, PPT, or PPTX creation and verification tasks.
Read, create, inspect, merge, split, rotate, encrypt, fill, and validate PDF files. Use when the user asks to work with a PDF or convert supported Markdown into a polished PDF in AstrBot.
Use this skill whenever the user mentions a PDF file or asks to produce/edit one. For read-only tasks such as reading, summarizing, extracting plain text, or answering questions from a PDF, follow this skill's read-only routing rules: use the built-in Read tool first, do not write code or scripts, and prefer…
docx
Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when…
Use this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting/decrypting PDFs, extracting images, and…