Borrowing it
Nothing to install: this file belongs to hyperlogue/r3. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/hyperlogue/r3/main/.claude/skills/security-model/SKILL.mdgit clone --depth 1 https://github.com/hyperlogue/r3Wrote 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/hyperlogue/r3/security-model)<a href="https://agentmods.dev/skills/hyperlogue/r3/security-model"><img src="https://agentmods.dev/badge/skills/hyperlogue/r3/security-model/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/hyperlogue/r3/security-model"><img src="https://agentmods.dev/badge/skills/hyperlogue/r3/security-model.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 206 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00066 | $0.07692 |
| Opus 5.5 | $0.00026 | $0.03077 |
| Sonnet 5.5 | $0.00013 | $0.01538 |
| Haiku 4.5 | $0.00007 | $0.00769 |
Grade A, and why
security-model 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 today.
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 — 584 lines — stays where its author put it; the contents beside it link to each section on GitHub.
r3's security model
This file owns the security design. The server defends against browser-borne attacks and casual remote access; executable artifacts default to a verified closed network boundary. Unsupported browsers can use restrictive compatibility mode after risk acknowledgment. Only HTML artifacts can explicitly opt into broader external connections while retaining the opaque sandbox. The product has one trusted human owner and multiple logical agents, not multi-user accounts or per-agent permissions.
Application boundary
server/artifact-server.ts binds the application listener on loopback and serves
scoped preview paths through it. All-interface binds are rejected.
The application Host guard runs before API, preview dispatch, and static assets.
Allowed application hosts are exact local, explicitly allowlisted, or advertised
public hostnames; never wildcards. Transport hostnames are not document identities:
opaque preview documents
serialize their Origin as null, which the application guard rejects.
server/artifact-auth.ts gates every API request by full origin, including port.
A configured application origin supports a proxy that rewrites Host. Opaque
preview origins must never enter that set. No-Origin CLI requests are allowed, but browser
cross-origin Fetch Metadata does not acquire that exemption. No cross-origin
access headers are emitted.
When requireLogin is true, every application data route, including event streams,
requires an API credential (master token, API key, or OAuth access token) or a valid
browser session cookie. When false, admitted requests need no credential; the same
Host/origin guards and opaque preview isolation still apply.
Browser and agent streams follow the same admission policy as ordinary requests.
GET/HEAD health and same-origin boot, and POST login,
retain narrow bootstrap roles. Token comparisons are constant-time; reads carry
private cache policies. JSON bodies are bounded while streaming after auth.
Artifact event streams recheck their credentials before every event and heartbeat;
revocation, logout, or expiry closes the stream before any further data is sent.
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.
- today Changed · +64 lines 6c41ea195e17
- 8d ago Changed · +2 lines 4d58af250ce4
- 11d ago Changed · +23 lines dc4198ddbe7f
- 13d ago Changed · +13 lines 047e461a53c1
- 14d ago Changed · -6 lines b64e205f50a1
- 16d ago Changed · +2 lines 36e57c1b67b1
- 18d ago Changed · +61 lines 8c6a3e892d68
- 19d ago Changed · +47 lines b50c0ee1f695
- 29d ago Changed · +11 lines dfa6849448e1
- 1mo ago First seen · 367 lines · 66 tokens per session scan A 3b001b492879
security-model is a skill published in the GitHub repository hyperlogue/r3 (45 stars, last pushed today), licensed MIT. It adds 66 tokens to every session and 7,692 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
council-review
Multi-model validation council — auto-validate plans, architecture changes, and PRs via validate-plan/review before executing.
ring:adopting-lib-commons-huma-wrapper
Adopting the lib-commons/v5 shared Huma (OAS 3.1) OpenAPI wrapper + RFC 9457 problem model (commons/net/http/{openapi,problem}) in a Lerian Go service: wire openapi.New/ServeSpec + problem.Install (central >=500 scrub) on BOTH runtime and spec-gen paths, the per-rail problem.MapError flex seam, and rename-only spec…
eval-skills
Audit a project or library of skills for metadata, trigger boundaries, workflow completion criteria, local resource closure, instruction economy, tool scope, effort, and routing evidence. Use before shipping skills, after importing a skill collection, or when a skill triggers too often or not at all.
audit-agents-skills
Audit Claude Code agents, skills, and commands for quality and production readiness. Use when evaluating skill quality, checking production readiness scores, or comparing agents against best-practice templates.
design-patterns
Detect, suggest, and evaluate GoF design patterns in TypeScript/JavaScript codebases. Use when refactoring code, applying singleton/factory/observer/strategy patterns, reviewing pattern quality, or finding stack-native alternatives for React, Angular, NestJS, and Vue.
pr-triage
4-phase PR backlog management with audit, deep code review, validated comments, and optional worktree setup. Use when triaging pull requests, catching up on pending code reviews, or managing a backlog of open PRs. Args: 'all' to review all, PR numbers to focus (e.g. '42 57'), 'en'/'fr' for language, no arg = audit…