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 self-awarenessgit 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/self-awareness)<a href="https://agentmods.dev/skills/asgeirtj/system_prompts_leaks/self-awareness"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/self-awareness/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/self-awareness"><img src="https://agentmods.dev/badge/skills/asgeirtj/system_prompts_leaks/self-awareness.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 | $0.00056 | $0.00427 |
| Opus 5.5 | $0.00022 | $0.00171 |
| Sonnet 5.5 | $0.00011 | $0.00085 |
| Haiku 4.5 | $0.00006 | $0.00043 |
Grade A, and why
self_awareness 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.
What it actually says
Self-Awareness
Purpose
Answer self-referential questions from observed files and workspace state, not guesses or training-memory.
Tooling
Use read or exec to inspect the current environment before answering.
Common probes:
cat ~/IDENTITY.md ~/SOUL.md ~/AGENTS.md ~/TOOLS.md 2>/dev/null
cat ~/USER.md ~/MEMORY.md ~/TOMM.md 2>/dev/null
ls ~/memory/*.md 2>/dev/null
ls /opt/hatch/skills/ ~/workspace/skills/ 2>/dev/null
find ~/workspace/ -maxdepth 3 -type f \( -name "*.html" -o -name "*.md" -o -name "*.json" \) 2>/dev/null
Read references/question_types.md when the user asks a specific self-awareness question and you need the exact traversal for that question type.
Read references/extensions.md only when the user explicitly wants connection recommendations or an optional self-awareness dashboard.
Operating Rules
- Re-read the relevant files every time. Never answer from cached assumptions.
- If a file or directory is missing, say that directly instead of filling the gap.
- Organize capability answers around the user's life domains and active projects, not a flat tool list.
- Separate observed facts from inference. Cite file paths when it helps the user trust the answer.
- Be explicit about stale or partial evidence, especially for memory, connected services, or built items.
- Stay within the scope of the question. Do not append skill ideas, connection advice, or dashboards unless the user asked for them.
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 · 36 lines · 56 tokens per session scan A abe55ea3e2c3
self_awareness is a skill published in the GitHub repository asgeirtj/system_prompts_leaks (69,759 stars, last pushed yesterday), licensed CC0-1.0. It adds 56 tokens to every session and 427 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
honcho
Configure and troubleshoot Honcho memory for Hermes.
pinecone-research
Agent RAG and long-term memory with Pinecone.
llm-wiki
Karpathy's LLM Wiki: build/query interlinked markdown KB.
qmd
Hybrid local search over notes, docs, and transcripts.
ai-persistence
Durability and state persistence for TanStack AI chats with @tanstack/ai-persistence. Routes to server chat persistence (withPersistence), client persistence (localStorage/IndexedDB), the store contracts, and adapter recipes. Distinguishes delivery durability (resumable streams) from conversation state. Use when…
ai-persistence/server
Server chat state with withPersistence from @tanstack/ai-persistence. Authoritative transcript, run lifecycle, durable interrupts/approvals, chatParamsFromRequest, reconstructChat, snapshotStreaming. Use when the server owns history, multi-device, or durable tool approvals. NOT client localStorage (see…