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 plaidgit 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/plaid)<a href="https://agentmods.dev/skills/asgeirtj/system_prompts_leaks/plaid"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/plaid/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/plaid"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/plaid.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 | $0.00028 | $0.01787 |
| Opus 5.5 | $0.00011 | $0.00715 |
| Sonnet 5.5 | $0.00006 | $0.00357 |
| Haiku 4.5 | $0.00003 | $0.00179 |
Grade A, and why
plaid 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 — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Plaid
Everything runs as plaid <command> and returns JSON for you to read, not to show the user. Add --help to any command to see its options.
Reads cover every linked institution at once. To narrow to one bank, pass --credential-id <id>, taking the id from body.institutions[] in an earlier read. Some reads also take --account-id <id> (repeatable) to narrow to specific accounts; when more than one institution is linked, pair it with --credential-id so the account ids resolve to the right bank.
Connecting
Plaid needs a one-time connect before any read returns data. Run plaid status. If it comes back not connected, post the exact connect_url it returns as [Connect Plaid](<connect_url>) and wait for the user to finish linking, then run plaid status again before reading. Don't invent a URL, send the user to Settings, or ask for bank credentials.
If the user asks to add or link another bank or institution, run plaid status and post the exact add_account_url it returns as [Add financial account](<add_account_url>). Never reuse connect_url for an additional institution. If add_account_url is absent, say that adding another institution is unavailable.
To disconnect every institution linked through Plaid, run plaid disconnect. If it returns a disconnect_url, post it as [Disconnect Plaid](<disconnect_url>) and wait for the user to confirm through that link; never claim disconnection before they confirm. If it has no URL, say Plaid is already disconnected. A bank-specific unlink is not available from chat: if the user asks to remove only one institution, direct them to that institution's account row in Settings instead of running the whole-Plaid disconnect. Don't set up a background check, goal, or reminder to monitor disconnection.
Common flows
Accounts and balances
List the linked accounts with plaid accounts — names, types, masked numbers, and each account's balances (balances carries available/current/limit). Start here to see what's linked and to answer "how much is in my checking," "what's my total across accounts," or net-worth questions. There is no separate balances command; plaid accounts is the balance read. Balances are Plaid's last reported figures, not live.
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.
- 12d ago First seen · 75 lines · 28 tokens per session scan A c1aa26261fd6
plaid is a skill published in the GitHub repository asgeirtj/system_prompts_leaks (69,759 stars, last pushed yesterday), licensed CC0-1.0. It adds 28 tokens to every session and 1,787 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 2026-10-05.
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.
evm
Read-only EVM client: wallets, tokens, gas across 8 chains.
hyperliquid
Hyperliquid market data, account history, trade review.
excel-author
Build auditable financial workbooks headless via openpyxl.