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 nebius/nebius-physical-ai --skill access-approvalgit clone --depth 1 https://github.com/nebius/nebius-physical-aiWrote 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/nebius/nebius-physical-ai/access-approval)<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/access-approval"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/access-approval/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/nebius/nebius-physical-ai/access-approval"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/access-approval.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00043 | $0.01000 |
| Opus 5 | $0.00022 | $0.00500 |
| Sonnet 5 | $0.00009 | $0.00200 |
| Haiku 4.5 | $0.00004 | $0.00100 |
Grade A, and why
access-approval 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 8d 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 — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Access Approval
Use NPA to discover gated HF/NGC dependencies, verify exact access, and preserve
a resumable handoff. Keep the human as the only party that can accept terms.
For the maintained readiness commands and credential resolution rules, read
skills/atomic/health-preflight/SKILL.md. For agreement scope, the Isaac-only
ACCEPT_EULA policy, and redistribution boundaries, read
skills/atomic/third-party-eula-preflight/SKILL.md.
Choose the narrow check
- During interactive
npa configure, offer the optional full-catalog audit. The explicit equivalent isnpa configure --prepare-catalog-access; opening pages remains a separate affirmative choice. Declining or postponing this audit must not block public assets or unrelated capabilities. - Immediately before workflow execution, resolve only the selected toolRefs'
exact HF/NGC dependency closure and require
Readybefore provisioning or runtime startup. Apply the same gate to local, hosted, serverless, VM, and Kubernetes execution; backend choice must not bypass artifact access. - Re-run the exact check after the human completes upstream steps, then resume with the persisted redacted command. Do not treat a prior full-catalog audit as permanent authorization.
Interpret evidence, not credentials:
Ready: an exact-revision HF payload-byte authorization probe or exact NGC registry artifact probe succeeded, or a cached success still matches the credential fingerprint, artifact revision, payload probe, and terms revision. It means technical fetch entitlement at that moment, never legal acceptance.Pending: credentials or a human-bound upstream approval are still needed.Denied: the provider rejected the credential or exact entitlement.Unavailable: the exact probe could not produce reliable evidence; do not convert this into success.
An HF token or NGC key proves identity only. Tokens inherit account access; they do not own licences. A generic login, HF repository/revision metadata response, or registry token exchange is not entitlement evidence. Probe a representative payload path for every exact model or dataset revision and the exact NGC container needed by the selected operation. Never select README, model-card, licence, tokenizer, or config-only files because those can remain public before approval.
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.
- 8d ago First seen · 91 lines · 43 tokens per session scan A ed231e9e9a21
access-approval is a skill published in the GitHub repository nebius/nebius-physical-ai (29 stars, last pushed today), licensed Apache-2.0. It adds 43 tokens to every session and 1,000 once invoked, about $0.0002 per session on Opus 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 2026-09-03.
Other skills, from other repositories
atmos-aws-compliance
AWS compliance commands in Atmos: atmos aws compliance report, Security Hub standards, CIS AWS, PCI DSS, SOC2, HIPAA, NIST, report formats, AI summaries.
x-scorecard
OpenSSF Scorecard for assessing open source project security. Check security best practices and compliance. Dependency: This is an x-cmd module. Install x-cmd first (see x-cmd skill for installation options). see x-cmd skill for installation.
Cloud Security & Container Hardening
AWS/Azure/GCP security auditing, container and Kubernetes hardening, Infrastructure as Code scanning, and cloud compliance assessment.
compliance-governance-audit
Audit Azure environment for compliance and governance posture. Checks Azure Policy, RBAC, tagging, resource locks, naming conventions, and regulatory alignment. Use when asked about compliance, governance, audit readiness, or policy violations.
alibabacloud-validate
Validate generated Terraform code against design requirements and quality standards. Dispatches spec compliance review and code quality review as parallel subagents. Remote syntax validation is owned by alibabacloud-terraform-codegen Step 6 and is NOT re-run here. WHEN: validate terraform, check code quality, review…
lov-domain-cutover
A workflow for putting a mainland-China website online after its ICP filing is approved. It verifies the deployment, DNS, HTTPS certificate, domain response, ICP footer, and handling of the old domain.