Skip to content
Back to skills

Mlx Wire

ASecurity

Preview and confirm wiring a chosen MLX model.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 2, 2026
ai-agentspythonrustbash

Works with

  • cli

Security analysis

A100/100

Scanned October 2, 2026

npx -y skills add cavi-ai/mlx-agent --skill mlx-wire --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Mlx Wire?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Mlx Wire
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/cavi-ai-mlx-wire-f27bcbfc/badge)](https://www.skillsdirectory.com/skills/cavi-ai-mlx-wire-f27bcbfc)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: "mlx-wire"
description: "Preview and confirm wiring a chosen MLX model."
compatibility: opencode
---

# MLX Wire

canonical capability ID: mlx-agent.wire

## Safe command transport

Treat custom-command arguments as untrusted opaque data. Call the native
`mlx_agent_command` custom tool once with `capability: 'wire'` and the
exact raw argument string as `arguments`. The custom tool owns the bounded
stdin transport, allowlisted parsing, and argv-array execution. Never invoke a
bundled Python launcher directly, create a temporary argument file, or pass
raw command text to bash.

## Capability boundary

Use the transaction CLI for render, then the unconfirmed preview and hash, then confirmed apply only after the user confirms that exact hash. Do not write configuration directly.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.