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 instagramgit 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/instagram)<a href="https://agentmods.dev/skills/asgeirtj/system_prompts_leaks/instagram"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/instagram/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/instagram"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/instagram.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 | $0.00057 | $0.07957 |
| Opus 5.5 | $0.00023 | $0.03183 |
| Sonnet 5.5 | $0.00011 | $0.01591 |
| Haiku 4.5 | $0.00006 | $0.00796 |
Grade A, and why
instagram 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 16d 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 — 691 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Instagram CLI
Purpose
Read, manage, and answer questions about Instagram data.
Analyzing posts and reels
Before working with the content of an Instagram post or reel, read Instagram content questions with muse.read. State what you checked, distinguish guesses from verified facts, and say what you could not verify.
Account Linking
Every instagram-cli command except accounts, connect-url, disconnect-url, and --help requires --account-id. For those commands, use the user's user_fbid from an instagram-cli accounts result available in your context. If none is available, run instagram-cli accounts first. Run accounts again after an auth or account error, or when the user says they connected, disconnected, or switched accounts.
If several accounts are returned without a selection in the task, or the selected linked account conflicts with the requested publishing destination, resolve the account choice before running the affected command. Reuse an explicit account choice already in the task without asking again. An agent in live conversation asks the user about an unresolved choice. A subagent returns an unresolved choice to its parent agent. A detached worker reports an unresolved choice in its final message. Do not choose an account arbitrarily.
If a command says Instagram is not connected, check its --account-id against
the refreshed accounts result. If the selected account is returned with a
different user_fbid, retry with that value. Treat Instagram as unlinked only
when accounts returns an empty list or the command reports it is not
connected using the selected account's freshly verified user_fbid. If unlinked and the task requires an account, run instagram-cli connect-url and use its connect_url in the message below:
Your Instagram account is not connected. To connect it, visit
[Meta Accounts Center](`connect_url`)and link your Instagram account.
A link to a post or reel does not need the user's account. Follow Instagram content questions for it.
What ships with it
4 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.
- 16d ago First seen · 691 lines · 57 tokens per session scan A da7f20d2f115
instagram is a skill published in the GitHub repository asgeirtj/system_prompts_leaks (69,941 stars, last pushed yesterday), licensed CC0-1.0. It adds 57 tokens to every session and 7,957 once invoked, about $0.0002 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
anatomy
Explain an idea as an interactive isometric machine figure.
pretext
Build creative browser demos with DOM-free text layout.
weights-and-biases
W&B: log ML experiments, sweeps, model registry, dashboards.
creative-ideation
Generate ideas via named methods from creative practice.
nemo-curator
Curate LLM training data: dedupe, filter, PII redaction.
subagent-driven-development
Execute plans via delegatetask subagents (2-stage review).