Local Studio is a local-first control panel for running, managing, and using self-hosted large language model backends such as vLLM, SGLang, llama.cpp, and ExLlama. It lets users launch models, monitor GPU and runtime state, chat through OpenAI-compatible endpoints, and run agent sessions from a desktop or connected mobile device. The catalogue skills and instruction support agent workflows with Local Studio and its controllers.
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 sybil-solutions/local-studio --skill cuagit clone --depth 1 https://github.com/sybil-solutions/local-studioWrote 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/sybil-solutions/local-studio/cua)<a href="https://agentmods.dev/skills/sybil-solutions/local-studio/cua"><img src="https://agentmods.dev/badge/skills/sybil-solutions/local-studio/cua.svg" alt="Measured on agentmods" 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 | $0.00054 | $0.00679 |
| Opus 5 | $0.00027 | $0.00340 |
| Sonnet 5 | $0.00011 | $0.00136 |
| Haiku 4.5 | $0.00005 | $0.00068 |
Grade A, and why
cua 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 7d 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.
How it starts
The opening of the file, as written. The whole thing — 35 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cua — the sandbox browser
The browser_* tools drive a headless Chromium-family browser that Local Studio launches on this machine — Chromium, Chrome, or Brave, whichever the user picked in the Browser panel. It has a throwaway profile: no saved logins, no cookies, no extensions, no downloads, one page at a time. Only public http(s) URLs and localhost are reachable. The user can watch it live in the Browser panel. When no browser binary is available the runtime degrades to a read-only fetch of the page rather than failing.
This is not the user's browser. If a page needs their session, this browser will show you the logged-out version of it and you will summarize the wrong thing.
When the task needs the user's own browser — their mail, a dashboard, anything behind a sign-in, or a tab they already have open — use the chrome_* tools instead. If those tools are not in your list, the user has not armed Chrome for this session: say so and tell them the composer's browser button turns it on. Never present a signed-out sandbox page as their account.
Tools
browser_navigate— open an absolutehttp(s)URL and wait for load.browser_get_url— where this browser actually is right now.browser_get_text— visible page text. The default way to read a page.browser_get_html— rendered HTML, when you need selectors or attributes.browser_screenshot— PNG data URI, when visual layout matters.browser_click/browser_fill— act on a CSS selector.browser_scroll— pixel delta, for lazy-loaded content.browser_back/browser_forward/browser_reload— history and refresh.browser_history— what this browser has already done and visited this session.
Protocol
- Navigate first, then read. Do not summarize a page you have not read this turn.
- Read with
browser_get_text. Reach forbrowser_get_htmlonly when text is not enough, andbrowser_screenshotonly when layout matters and the model has vision. - Before repeating work, call
browser_history— it also shows pages the user opened in the panel, so it is the honest answer to "where are we". - A
found: falsefrom click or fill means the selector was wrong. Re-read the page and pick a new selector instead of retrying the same one. - Experiment freely. This browser is signed out and unwatched, so a wrong click costs a reload — that is the whole point of having it separate from
chrome_*. - Never type credentials or payment details here. There is no session to sign into, so it achieves nothing and only exposes the secret.
- If a tool reports the browser is unavailable, say so in one line. Never claim to have opened or inspected a page you could not reach.
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.
- 7d ago First seen · 35 lines · 54 tokens per session scan A a32e7abcb206
cua is a skill published in the GitHub repository sybil-solutions/local-studio (1,752 stars, last pushed 4d ago), licensed Apache-2.0. It adds 54 tokens to every session and 679 once invoked, about $0.0003 per session on Opus 5. 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 skills, from other repositories
nyx-ui-screenshots
Regenerate the README UI screenshots for nyx-local-ai from the browser harness — build the webview, serve the harness scenes, capture and crop the images into docs/. Use when the UI changed visibly, when the user asks to update screenshots, or before a release with UI changes.
mood-tracker
A simple mood tracking skill that stores your daily mood and comments. Use this when the user wants to log their mood, track how they feel, or see their mood history.
translator
Translates text between languages. Use this when the user wants to translate a word, phrase, or passage into another language.
mood-music
A skill to suggest or play music based on the user's mood, including analyzing images or audio, by querying available genres and generating music via the Loudly API.
budget-tracker
Track daily expenses and income. Use when the user wants to log spending, record a purchase, add income, or see a budget summary.
kitchen-adventure
Act as a dungeon master for a text-based adventure set in a world where everyone is a sentient kitchen appliance. Trigger when user says "start kitchen adventure".