start-stop

Commands for controlling persistent sandboxes, which are isolated environments that can be stopped and started later.

In plain words
What is it for?
Use them to start, stop, pause, resume, or fork sandboxes. Some Firecracker sandboxes can preserve memory and running-process state.
Why use it?
They let you manage an environment across its running, stopped, paused, and resumed states.

Command

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/thrashr888/agentkernel/start-stop
Clone the repo
git clone --depth 1 https://github.com/thrashr888/agentkernel
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,557 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00000 $0.01557
Opus 5 $0.00000 $0.00779
Sonnet 5 $0.00000 $0.00311
Haiku 4.5 $0.00000 $0.00156

Measured 2d ago against content hash 377560973505, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

start-stop 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 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.

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.

docs/commands/start-stop.md · 245 lines

How it starts

The opening of the file, as written. The whole thing — 245 lines — stays where its author put it; the contents beside it link to each section on GitHub.

agentkernel sandbox start / stop / pause / resume / fork

Manage the lifecycle of persistent sandboxes.

start and stop are the portable lifecycle operations. Firecracker sandboxes on x86_64 Linux/KVM also support full-state pause, resume, and fork, which preserve guest memory and process state.

start

Start a stopped sandbox.

Firecracker sandboxes are started through the long-running local API server so the VM process survives after the CLI exits. Run agentkernel serve first; AGENTKERNEL_PORT selects a non-default local port. If the server requires an API key, set AGENTKERNEL_API_KEY for delegated CLI requests.

Each ordinary Firecracker start owns an AgentKernel-managed writable rootfs lineage. The opaque lineage reference is stored with the sandbox state and is reopened after stop, so guest filesystem changes survive a cold start.

Set AGENTKERNEL_ROOTFS_COW_MAX_BYTES to a host-wide logical-byte cap for durable Firecracker writable images. A new start fails before copying its base image when existing COW images plus that image would exceed the cap. This is a storage headroom guard, not a tenant quota; it applies to every local Firecracker sandbox and counts stopped lineages as well as running ones.

Before delegating, the CLI resolves the saved agentkernel.toml permissions and startup file injections. It stores that resolved configuration in a private host manifest and sends only an unguessable one-shot reference in the local start request. The manifest is bound to the sandbox UUID and owner, consumed on the first attempt, expires after five minutes, and is scrubbed on removal. A valid reference can authorize the server's first claim of an unowned local sandbox even after the server restarted or previously inspected it. This preserves the selected profile, network and mount behavior, resource limits, and file contents while the server retains VM ownership, without exposing a public capability-escalation payload.

Usage

agentkernel sandbox start [OPTIONS] <NAME>

Read the full file on GitHub · 245 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. 2d ago First seen · 245 lines · 0 tokens per session scan A 377560973505

Subscribe to this mod's changes

start-stop is a command published in the GitHub repository thrashr888/agentkernel (58 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,557 tokens. 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.