Borrowing it
Nothing to install: this file belongs to opspresso/vibemon-app. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/opspresso/vibemon-app/main/.claude/skills/run-desktop/SKILL.mdgit clone --depth 1 https://github.com/opspresso/vibemon-appWrote 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/opspresso/vibemon-app/run-desktop)<a href="https://agentmods.dev/skills/opspresso/vibemon-app/run-desktop"><img src="https://agentmods.dev/badge/skills/opspresso/vibemon-app/run-desktop/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/skills/opspresso/vibemon-app/run-desktop"><img src="https://agentmods.dev/badge/skills/opspresso/vibemon-app/run-desktop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 61 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
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 | $0.00069 | $0.01449 |
| Opus 5.5 | $0.00028 | $0.00580 |
| Sonnet 5.5 | $0.00014 | $0.00290 |
| Haiku 4.5 | $0.00007 | $0.00145 |
Grade A, and why
run-desktop scanned grade A 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 15d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -X POST http://127.0.0.1:19280/status -H 'Content-Type: application/json' \ How it starts
The opening of the file, as written. The whole thing — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
VibeMon starts in the tray. A local POST /status or fresh selected cloud-agent
activity creates its character window on demand. For automated regression
checks, use pnpm test:overlays and the isolated
workspace fixture with pnpm test:workspace. For interactive
agent verification, run pnpm preview:workspace and use the UI tools provided by the current session. The preview uses synthetic data, a loopback API, and fake integration/credential ports. It does not read or change the user's hooks or keychain.
The optional driver.mjs is a manual CDP utility; do not use it when session
instructions require the provided computer-use tools.
The manual driver isolates the Electron store with --user-data-dir. It
launches the full application, which still manages ~/.vibemon/config.json
and refreshes local provider usage. Use the native fixtures when shared hook
configuration must remain untouched.
Optional manual driver
One shot: pass the commands as arguments. They run in order, the app is closed at the end, and the exit code is non-zero if any command failed.
node .claude/skills/run-desktop/driver.mjs \
"launch 3d" "lock codex" "status working" "engine" "ss codex-3d" "logs"
Then open the PNG and look at it. Screenshots land in /tmp/vibemon-shots/
(override with SCREENSHOT_DIR).
Run it with no arguments for an interactive driver> REPL instead (same
commands, quit to exit).
The real app must not be running — it owns port 19280, and launch fails fast
with a clear error if the port is taken.
Commands
| command | what it does |
|---|---|
launch [2d|3d] |
seed the render mode, boot the app, wait for its HTTP server |
status <state> [character] [project] |
POST /status → creates/updates the character window, waits for it |
lock <character|auto> |
POST /character-lock — pins the character against other bridges |
focus |
what the window is actually showing (focused project / character / state) |
ss [name] |
screenshot over an opaque backdrop, prints focus alongside |
ss-raw [name] |
screenshot the true transparent frame |
engine |
which engine booted: vibemon-canvas (2D) vs vibemon-canvas-3d (WebGL) |
page <character|bubble> |
retarget commands at the character window or the speech bubble |
windows |
list open windows |
eval <js> / text [sel] |
evaluate in the page / dump innerText |
wait <ms> |
pause the sequence (e.g. to let an animation settle) |
logs |
renderer console, uncaught errors, CSP violations, and main-process output since launch |
quit |
close the app (implicit at the end of a one-shot run) |
What ships with it
1 file beside SKILL.md 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.
- 15d ago Changed e3b9c78a9d66
- 18d ago Changed 7e502417fa7c
- 1mo ago First seen · 110 lines · 69 tokens per session scan A 25714a56fb87
run-desktop is a skill published in the GitHub repository opspresso/vibemon-app (23 stars, last pushed today), licensed MIT. It adds 69 tokens to every session and 1,449 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
reversi-master
Manage ChatGPT, Codex CLI, and Claude Code accounts and track quota in one place. Use multi-model smart routing with priority and automatic failover. Install AI tools and deploy local LLMs in one click.
Trash Talk
AI becomes provocative and mocks the opponent during gameplay.
openwiki-operator
Use when working with OpenWiki or configured OpenWiki MCP tools to search, add sources, propose pages or edits, validate proposals, review/apply changes, monitor agent activity, or explain wiki Git history.
autoresearch
Run an auto research / autoresearch loop for measurable improvements. Use when optimizing a skill, agent, prompt, code path, benchmark, or other target with a mechanical metric: baseline, mutate one thing, verify, keep or discard, log, chart progress, and optionally save improved custom skills or agents through Open…
openwiki-meeting-curation
Use when turning transcript evidence into OpenWiki meeting pages, entity updates, decisions, action items, and proposal-safe wiki changes.
gateway-review-gate
Spec-driven review and verification gate for Gateway tasks across code, docs, slides, research, operations, and other deliverables.