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 robium-ai/robium --skill live-demogit clone --depth 1 https://github.com/robium-ai/robiumWrote 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/robium-ai/robium/live-demo)<a href="https://agentmods.dev/skills/robium-ai/robium/live-demo"><img src="https://agentmods.dev/badge/skills/robium-ai/robium/live-demo.svg" alt="Measured on agentmods" 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 | $0.00015 | $0.00737 |
| Opus 5 | $0.00008 | $0.00368 |
| Sonnet 5 | $0.00003 | $0.00147 |
| Haiku 4.5 | $0.00002 | $0.00074 |
Grade A, and why
live-demo 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 — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Live demo
Separate the host-level control plane from the visitor's session data plane.
Start from a working app
- Add hosted lifecycle only after the application's normal smoke test passes.
- Define one real demo outcome, app-specific readiness, maximum session time, fleet capacity, cost boundary, and authoritative teardown.
- Keep the app's runtime and
demo.orchestratorfacts inrobium-app.yaml. The website/orchestrator derives its registry entry from that source; current production enablement remains website-owned. - Never allocate on catalog, overview, or idle live-page load. Starting paid capacity requires an explicit visitor action.
- Keep public project copy, cards, articles, and media in
app-publishing.
Keep the session boundary explicit
- A host-level lifecycle layer enforces fleet limits and expiry and returns an instance ID plus connect host. Local Docker and RunPod allocate explicit resources; Cloud Run routes requests to a predeployed service.
- The verified gateway example gives the first active claimant exclusive use and rejects concurrent foreign claims. Its browser-provided session ID is a coordination key, not authentication.
- When access control matters, issue a signed or high-entropy, host-bound capability outside the instance and validate it on every lifecycle and viewer route. Affinity and hard-to-guess URLs remain routing conveniences.
- Keep the orchestrator out of the viewer data path. The browser should talk directly to the gateway after allocation.
- Read orchestrator-pattern.md when implementing provider drivers or local start/delete behavior. Read gateway-pattern.md when implementing claims, viewer transport, readiness, or shutdown.
Make lifecycle visible
- Expose idle, allocating, booting, ready, reconnecting, stopping, expired, busy, budget-exhausted, and failed states honestly.
- Keep the current instance while reconnecting; a transient poll failure must not allocate a replacement.
- Show real boot messages, remaining time, capacity, and expected cold-start behavior. Avoid invented percentage progress.
- Read session-ui.md for the shared browser state machine. Read viewers.md when choosing or embedding Gradio, Lichtblick, Foxglove, or a custom interaction surface.
What ships with it
7 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.
- today Changed · -142 lines · -158 tokens per session 671f7dd46a7e
- 5d ago First seen · 214 lines · 173 tokens per session scan A 734253f93d78
live-demo is a skill published in the GitHub repository robium-ai/robium (9 stars, last pushed today), licensed MIT. It adds 15 tokens to every session and 737 once invoked, about $0.0001 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
together-kueue
Install and use the Kueue job-queueing controller on a Together AI Kubernetes GPU cluster to gate jobs on quota. Covers installing Kueue, defining ResourceFlavor, ClusterQueue, and LocalQueue quota, submitting jobs to a queue, and watching quota admit or suspend them. Reach for it when a Together cluster's GPU pool…
edge-computing-expert
Design and implement production-grade edge computing solutions including CDN optimization, fog computing architectures, and distributed edge workloads. Use when the user mentions edge computing, CDN or edge caching, fog computing, MEC, IoT edge workloads, or latency-sensitive distributed deployments.
Infrastructure Development
Enterprise infrastructure patterns for AWS VPC, ECS, and PrivateLink deployments.
mk8s-byok
Runs workloads on your own hardware and provisions managed Kubernetes (mk8s) clusters on Control Plane. Use when the user asks about bare metal, on-prem, data centers, their own servers, mk8s, BYOK, or node pools.
proxmox-engineer
Proxmox VE e PBS — cluster, quorum, Ceph, LXC, VM, HA, replicação, storage e backup. Use quando o usuário disser "proxmox", "PVE", "ceph", "LXC", "container LXC", "VM não sobe", "cluster sem quorum", "PBS", "backup da VM". Sem MCP nem CLI padrão instalada — opera via SSH no nó ou API REST.
graviton-migration
Execute a workload migration from x86 (amd64) to AWS Graviton (arm64) — set up the Arm migration MCP server for code, dependency, and container arm64-readiness scanning, then run the migration — pre-migration scanning, Karpenter arm64 node cutover, and multi-arch CI pipelines. Use when someone says "migrate to…