agentmods.dev

launch

launch is a command for coding agents from vast-ai/vast-claude-plugin. It costs 26 tokens per session (1,577 once invoked), scanned C, original, MIT.

A shortcut for launching a rented Vast.ai GPU computer from an offer ID with a PyTorch-based image and connection settings suited to SSH access.

In plain words
What is it for?
Start a Vast.ai instance with optional disk size, label, image, or interruptible pricing from a single offer ID.
Why use it?
It handles required SSH-key checks and launch options so the new computer is reachable and unavailable offers are cancelled automatically.

Command

Part of the vastai plugin — 2 skills, 5 commands shipped together

not rated 3 2mo ago C scan 26 tokens original MIT
View the source file vast-ai/vast-claude-plugin
Install

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.

agentmods
npx agentmods add commands/vast-ai/vast-claude-plugin/launch
Clone the repo
git clone --depth 1 https://github.com/vast-ai/vast-claude-plugin

Or install vastai, the plugin that ships this one along with the rest of its 2 skills, 5 commands.

Wrote 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.

agentmods badge for launch

README.md
[![agentmods](https://agentmods.dev/badge/commands/vast-ai/vast-claude-plugin/launch.svg)](https://agentmods.dev/commands/vast-ai/vast-claude-plugin/launch)
Your own site
<a href="https://agentmods.dev/commands/vast-ai/vast-claude-plugin/launch"><img src="https://agentmods.dev/badge/commands/vast-ai/vast-claude-plugin/launch.svg" alt="Measured on agentmods" height="20"></a>
Per session 26 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,577 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00026 $0.01577
Opus 5 $0.00013 $0.00788
Sonnet 5 $0.00005 $0.00315
Haiku 4.5 $0.00003 $0.00158

Measured 4d ago against content hash 9896c26acada, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

launch scanned grade C 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 4d 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.

Reaches for credential fileshighPrivilege escalation

SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.

vastai create ssh-key "$(cat ~/.ssh/id_ed25519.pub)" --raw
commands/launch.md · 83 lines

How it starts

The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/vastai:launch

Launch a Vast.ai GPU instance from an offer ID. Defaults to the Vast-curated PyTorch image and the --ssh --direct --cancel-unavail connection mode — what most renters want.

Steps

  1. Parse $ARGUMENTS as: <offer-id> [image] [--disk N] [--label STR] [--bid PRICE]. The offer ID is required; everything else has a default.

  2. Ensure an SSH key is registered BEFORE launching. Launching without a registered key produces an unreachable instance.

    vastai show ssh-keys --raw

    If the response is [], register the user's default key first (public key is POSITIONAL — there is no --ssh-key flag):

    vastai create ssh-key "$(cat ~/.ssh/id_ed25519.pub)" --raw

    Fall back to ~/.ssh/id_rsa.pub if no ed25519 key exists; instruct the user to generate one if neither file exists.

  3. Honor pricing-mode intent from the arguments. vastai search offers does not support filtering by id=, so there is no clean way to "look up the offer" before launching — accept the user's stated intent:

    • If $ARGUMENTS includes --bid PRICE, the user wants interruptible pricing — pass --bid_price PRICE to create instance (step 5).
    • Otherwise, launch on-demand. Every offer can be rented either way; the on-demand price is the offer's dph_base and the bid floor is its min_bid. The user picks the mode at launch.

  4. Build the create command with these defaults:

    • --image → user-provided, else vastai/pytorch:@vastai-automatic-tag (Vast-curated image; @vastai-automatic-tag only resolves on vastai/* images, NOT pytorch/pytorch or other third-party images)
    • --disk → user-provided, else 20
    • --label → user-provided, else claude-launch-$(date +%s) (so the instance is greppable later)
    • Always include --ssh --direct --cancel-unavail --raw. --cancel-unavail fails the launch if the machine becomes unavailable mid-call instead of silently producing a stopped instance that accrues disk charges while you poll forever.
    • If a bid price was supplied, also add --bid_price <PRICE>. (No need to pass --type bid on create instance — --bid_price selects bid mode.)

Read the full file on GitHub · 83 lines

Changes

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.

  1. 4d ago First seen · 83 lines · 26 tokens per session scan C 9896c26acada

Subscribe to this mod's changes

launch is a command published in the GitHub repository vast-ai/vast-claude-plugin (3 stars, last pushed 2mo ago), licensed MIT. It adds 26 tokens to every session and 1,577 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 1 finding (reaches for credential files). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

CloudDevOps and cloud
Related

Other commands, from other repositories

deploy

Build, test, deploy with staged rollout.

dinhnguyenngoc/spec-driven-claude-code · 10 tokens

deploy

Deploy a frontend (React, Next.js, or static HTML) to a live URL on Butterbase.

butterbase-ai/butterbase-skills · 20 tokens

cf-scaffold-project

Scaffold a Cloudflare project — Worker, Pages, or Worker+D1+R2 starter — with Wrangler config, Terraform skeleton, and GitHub Actions deploy using scoped API tokens.

jsgerman-oss/blackrim-nimbus-skills · 37 tokens

scan

Scan AWS account for cost optimization.

prajapatimehul/claude-aws-cost-saver · 6 tokens

finops-feedback

Step 5 (Feedback Loop & Celebration) — measure realized against projected savings, compute a labelled Cloud Entropy proxy, close the opportunity, and emit at least one new idea or policy update so the loop actually closes. Applies the double-loop gate. Mutates on the closure path.

glassity/agent-skills · 57 tokens

run-panguweather

Guide the user through running PanguWeather end-to-end on an AMD cluster.

AMDResearch/ai4science-studio · 0 tokens