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 slidesgit 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/slides)<a href="https://agentmods.dev/skills/asgeirtj/system_prompts_leaks/slides"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/slides/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/slides"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/slides.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 | $0.00017 | $0.04486 |
| Opus 5.5 | $0.00007 | $0.01794 |
| Sonnet 5.5 | $0.00003 | $0.00897 |
| Haiku 4.5 | $0.00002 | $0.00449 |
Grade A, and why
slides 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 10d 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 — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Slides
This artifact is one release of the Slides runtime: read-only, the type's. A deck's content is ITS
OWN files under project/; write only there. A deck inherits the type's capabilities
{"downloads":{},"artifact":{},"comments":{"composer_only":true,"customAnchors":true},"room":{},"db":{"rules":[{"path":"","write":"admin"},{"path":"notes","read":"admin","write":"admin"}]},"assets":{},"user":{"scopes":["profile"]},"mcp":{"servers":[{"server":"Google Drive","tools":["create_file","gdrive_upload"]}]}}
and its contract "0.2.47".
The deck exists
Work on THAT deck, its url on every call: never create
another or send type_url again. read project/deck.json: none, or
its order is empty: an empty deck, see Creating. Else see Revising.
If no deck exists yet: one call with type_url = the Slides type's link and
title = the deck's name (REQUIRED),
auto_open: "after_first_write" if offered, and no files; later calls use
the reply's url.
Tell the user what happens to the deck, never the mechanism.
Asked to send it to apps like Canva: use the link Send to just posted, else a tool that makes one, else point the user to Share › Export › Send to, if shown, which posts the link here. Give an app's import tool no other link, never the deck's url.
The deck's files
project/deck.json, the index:
{"v":4,"createdOnFiles":{"v":1,"at":"2026-09-14T18:20:00Z"},"lists":"css","title":"Q3","order":["cover","plan"],"sections":{"s1":{"description":"Numbers","start":"cover"}},"faces":{"source-serif-4":{"family":"Source Serif 4","href":"https://fonts.googleapis.com/css2?family=Source+Serif+4:[email protected]&display=swap"}},"designSystems":[]}.
createdOnFilesandlists: an index YOU create has both exactly so (nolistsif the deck already had slides),at= now.order= slide ids in deck order.sections= your outline: any key → a run's one sentence and its first slide's id; the first starts at the cover.cover= the cover slide's id. Ids:[A-Za-z0-9_-]{1,64}.faces= one entry per typeface the slides name (at most 4), keyed by the family lowercased, spaces as-:family(a letter first, then letters, digits, spaces,_,-; 40 at most) plus EITHERhref(only ahttps://fonts.googleapis.com/css2?…link) ORsrc:"/_blob/<id>"(an uploaded woff2/woff/ttf/otf, step 2) or"project/ds/<folder>/<path>"(step 3); a rule broken: the entry is ignored. Basic/generic faces need no entry.
designSystemsandproject/ds/<folder>/: written only by the install (step 3). An existing index: keep every key you don't change, named here or not; never addlists.project/slides/<id>.html, one per slide: EXACTLY ONE
What ships with it
12 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.
- artifact-type/reference/craft.md 11 KB
- artifact-type/reference/deck-files.md 16 KB
- artifact-type/reference/diagram-recipes.md 14 KB
- artifact-type/reference/diagrams.md 16 KB
- artifact-type/reference/fonts.md 12 KB
- artifact-type/reference/format.md 16 KB
- artifact-type/reference/images.md 7.9 KB
- artifact-type/reference/layout.md 11 KB
- artifact-type/reference/photo-library.md 3.3 KB
- artifact-type/reference/questions.md 4.0 KB
- artifact-type/reference/styles.md 7.8 KB
- artifact-type/reference/view-state.md 8.6 KB
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.
- 10d ago First seen · 155 lines · 17 tokens per session scan A bc23f87a584e
slides is a skill published in the GitHub repository asgeirtj/system_prompts_leaks (69,918 stars, last pushed yesterday), licensed CC0-1.0. It adds 17 tokens to every session and 4,486 once invoked, about $0.0001 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 2026-10-10.
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.
siyuan
Query and edit a SiYuan knowledge base via its API.