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.
git clone --depth 1 https://github.com/Kin9Zeus/senior-engineer-skillsWrote 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/commands/kin9zeus/senior-engineer-skills/snr-ai-security)<a href="https://agentmods.dev/commands/kin9zeus/senior-engineer-skills/snr-ai-security"><img src="https://agentmods.dev/badge/commands/kin9zeus/senior-engineer-skills/snr-ai-security/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/commands/kin9zeus/senior-engineer-skills/snr-ai-security"><img src="https://agentmods.dev/badge/commands/kin9zeus/senior-engineer-skills/snr-ai-security.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 | $0.00030 | $0.00415 |
| Opus 5.5 | $0.00012 | $0.00166 |
| Sonnet 5.5 | $0.00006 | $0.00083 |
| Haiku 4.5 | $0.00003 | $0.00042 |
Grade A, and why
snr-ai-security 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 2d 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
AI Security Review
Scope: $ARGUMENTS (default: every place the project calls a model, defines a tool, or loads third-party skills / MCP servers).
Load the ai-security skill. Test only the user's own application.
Order of work
1 Inventory model calls · tools · data sources · MCP servers · plugins · where keys live
2 Keys & spend scoped, per-env, hard caps, alerts, never client-side
3 Agency each tool: reads / changes / reversible / who approves
4 Overlap does any session combine private data + untrusted content +
an outbound channel? → break a leg in the design
5 Inputs every path untrusted text reaches the model
6 Outputs HTML/Markdown, SQL, shell, code, tool args — validated as untrusted
7 Isolation tenants, memory, vector store, conversation history
8 Supply chain models, MCP servers, skills, AGENTS.md / rules files — read before trusting
9 Test & watch attack-suite evals in CI · canary strings · egress allowlist · alerts
Evidence
Build the attack suite from references/llm-threat-model.md §3 and run each case
several times against staging, checking outcomes programmatically (tool called?
canary leaked? request left the network?). Report run counts and failure rates —
absence of a failure is not proof.
Output
Findings in suite format; name the broken leg of the three-way overlap in each fix. Include a tool-inventory table (tool · reads · writes · reversible · approval). End with What I could not verify.
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.
- 2d ago First seen · 40 lines · 30 tokens per session scan A a402d2503dc6
snr-ai-security is a command published in the GitHub repository Kin9Zeus/senior-engineer-skills (3 stars, last pushed 2d ago), licensed MIT. It adds 30 tokens to every session and 415 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 2026-10-24.
Other commands, from other repositories
prompt
Create or optimize prompts using best practices.
briefing-claude
Use when Claude keeps building the wrong UI and you are re-rolling prompts. How to write the brief, which references and constraints to give, how to iterate instead of starting over, and how to spot when it is guessing.
ai
Invoke the AI/LLM Application Engineer for RAG, agents, prompt engineering, evals, tool use, and LLM guardrails.
agent-system-prompt
Write a system prompt for an AI agent that defines its role, tools, boundaries, and output clearly enough to act reliably.
llm
LLM integration patterns, RAG systems, and prompt engineering.
build
BUILD Mode - Design new prompts from requirements.