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.
git clone --depth 1 https://github.com/milind-soni/axstreamWrote 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/plugins/milind-soni/axstream/axstream)<a href="https://agentmods.dev/plugins/milind-soni/axstream/axstream"><img src="https://agentmods.dev/badge/plugins/milind-soni/axstream/axstream/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/plugins/milind-soni/axstream/axstream"><img src="https://agentmods.dev/badge/plugins/milind-soni/axstream/axstream.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
axstream 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 1mo 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.
What it actually says
{
"name": "axstream",
"description": "Fast, verified computer use for coding agents: read the screen as text (~200ms), execute batched verified actions, and compile any task done once into a macro that replays in seconds with zero model calls.",
"version": "0.3.2",
"author": {
"name": "Milind Soni",
"url": "https://github.com/milind-soni"
},
"homepage": "https://axstream.dev",
"repository": "https://github.com/milind-soni/axstream",
"license": "MIT",
"keywords": [
"computer-use",
"automation",
"macos",
"mcp",
"macros"
]
}What it installs
The manifest is a name and a version. 1 skill, 3 commands, 1 MCP server travel with it, and installing the plugin installs all of them — 143 tokens a session between them. Each is measured on its own page, and each can be installed alone.
What ships with it
1 file beside plugin.json 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.
- 1mo ago First seen · 19 lines scan A a04b5588e16b
axstream is a plugin published in the GitHub repository milind-soni/axstream (20 stars, last pushed 2mo ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. 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.
Other plugins, from other repositories
background-computer-use
Drive macOS apps from your MCP-capable LLM client in the background: click, type, scroll, and inspect accessibility trees without activating apps or moving your cursor. Parity port of the Codex computer-use tool surface plus a small set of app-specific skills.
planchette
Point at any macOS window and let Claude Code control it — click, type, screenshot, and drive another app's window.
screenhand
Desktop and browser automation for Claude Code. Control any macOS/Windows app, automate social media, run QA tests, edit video, design in Figma/Canva, scrape the web, and orchestrate multi-agent workflows — 111 MCP tools.
clawdcursor
Safe cross-OS desktop & browser control for Claude Code — the fallback execution layer for when APIs, CLIs, and direct integrations aren't available. Registers the clawdcursor MCP server in compact mode and bundles its usage skill (the SKILL.md at the repo root). Requires Node.js 20+; the MCP server launches via npx…
kwin-mcp
MCP server + skill for Linux KDE Plasma 6 Wayland GUI automation. Provides 29 MCP tools (sessions, apps, screen capture, AT-SPI2 elements with refs, windows and virtual desktops, pointer, keyboard, touch, clipboard, D-Bus and Wayland escape hatches) plus the kwin-desktop-automation skill that guides session-mode…
browser-gateway
Cloud browsers through the hosted BrowserGateway MCP server. When WebFetch is blocked or gets an empty page, Claude retries with a real browser.