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 smixs/iva-agent --skill install-ivagit clone --depth 1 https://github.com/smixs/iva-agentWrote 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/smixs/iva-agent/install-iva)<a href="https://agentmods.dev/skills/smixs/iva-agent/install-iva"><img src="https://agentmods.dev/badge/skills/smixs/iva-agent/install-iva/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/smixs/iva-agent/install-iva"><img src="https://agentmods.dev/badge/skills/smixs/iva-agent/install-iva.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 | $0.00167 | $0.01205 |
| Opus 5.5 | $0.00067 | $0.00482 |
| Sonnet 5.5 | $0.00033 | $0.00241 |
| Haiku 4.5 | $0.00017 | $0.00120 |
Grade B, and why
install-iva scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
| `echo "$PASS" \| sudo -S` inside `ssh host "…"` | The password lands in the remote command line; see `references/secrets.md` | How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Install Iva
Overview
Take a person from "I have a server" (or "use this Mac") to "my Iva bot answers me", doing every technical step yourself. The person is usually not technical. Every run is repeatable: check the state first, do only what is missing, and leave a server that only the person's key can enter.
Talk to the person in their language, in plain words, one short message per step. Show them what you found before you change anything.
The ladder
Walk these steps in order on every run. Each step: check → do only what is missing → check again. A step that is already done costs one check, not a redo.
| Step | What it settles | Read before doing it |
|---|---|---|
| 0. Pre-flight | the person's computer can reach and manage a server; the server is fit for Iva | references/preflight.md |
| 1. Target | which machine: a server (address from env, ~/.ssh/config, mail, or the hosting email) or this Mac |
references/target.md |
| 2. Access | a key-only SSH login as user iva, saved as ssh iva on the person's computer |
references/access.md |
| 3. Server security | non-root user, key-only SSH, root login closed, firewall, fail2ban, security updates | references/server-security.md |
| 4. Iva | install, repair, update or leave alone — decided by what is already there | references/iva-state.md |
| 5. Keys | model, bot token, the person's Telegram ID — only what is missing | references/keys.md |
| 6. Finish | the bot answers the person; they know the one command to reach the server | references/finish.md |
A Mac target skips steps 2 and 3 and installs locally (same installer; it sets up launchd itself).
Rules that protect the person
- Secrets never become text. A password, bot token or API key goes from a file to the server through a pipe or a
file with mode 600 — never inside a command line (it shows in
pson both machines), never echoed back, never in your final message. Why: the session transcript, shell history and process list all outlive the install. Details and safe patterns →references/secrets.md. - Never lock the person out. Prove the new key login works before closing password or root login, and keep the current session open until then. Why: a VPS with closed password login and a broken key is only recoverable through the hosting provider's rescue console, which a beginner cannot use.
- Use the official installer.
bootstrap.shprepares the server,install.shinstalls Iva. Do not hand-write.envbeyond the minimal file inreferences/keys.md, do not write systemd units or sudoers files yourself. Why: hand-made setups break on the nextiva update, and the person cannot get help for a setup nobody else has. - Change only what is Iva's. On the person's computer touch only
~/.ssh/iva_ed25519*and theHost ivablock in~/.ssh/config. On a server that already runs other services, ask before installing next to them. - A question you cannot ask is a stop, not a guess. If you run without a way to ask (non-interactive mode) and a required answer was not given, stop with that one question.
What ships with it
8 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.
- 2d ago First seen · 68 lines · 167 tokens per session scan B 258b15a61f14
install-iva is a skill published in the GitHub repository smixs/iva-agent (263 stars, last pushed today), licensed MIT. It adds 167 tokens to every session and 1,205 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-19.
Other skills, from other repositories
openclaw-bun-upgrade
Qualify and promote a pinned Bun fork build for a Linux OpenClaw service, including upstream sync, release proof, native workload admission, and rollback planning.
architecture-diagram
Generate dark-themed SVG diagrams of software systems and cloud infrastructure as standalone HTML files with inline SVG graphics. Semantic component colors (cyan=frontend, emerald=backend, violet=database, amber=cloud/AWS, rose=security, orange=message bus), JetBrains Mono font, grid background. Best suited for…
serving-llms-vllm
Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production LLM APIs, optimizing inference latency/throughput, or serving models with limited GPU memory. Supports OpenAI-compatible endpoints, quantization (GPTQ/AWQ/FP8), and tensor parallelism.
ssh-remote
Manage SSH connections, key management, remote command execution, tunneling, and file transfer with scp and rsync. Use when the user asks to connect to a remote server, generate SSH keys, set up port forwarding or tunnels, transfer files remotely, configure jump hosts, manage knownhosts, or set up SSH config for…
stripe-projects
Provision SaaS services + sync creds via Stripe Projects.
lambdasim
Run the studio's per-project NLP Lambda path locally with lambdasim, haven's AWS Lambda stand-in: functions, invokes and response streams answered on this stack's nlpgo. Use when someone says 'haven lambda', 'lambdasim', 'haven up +lambda', 'haven sim lambda', 'NLP Lambda locally', 'per-project Lambda'…